2017 Edition TABLE OF CONTENTS 2017

Greeting Letter From the President and From the Chair 2

Flatiron Institute Flatiron Institute Inaugural Celebration 4

Center for Computational Quantum Physics 6

Center for Computational Astrophysics: Neutron Star Mergers 8

Center for Computational Biology: Neuronal Movies 10

Center for Computational Biology: HumanBase 12

MPS Scott Aaronson: Quantum and Classical Uncertainty 14 Mathematics and Physical Sciences Sharon Glotzer: Order From Uncertainty 16

Horng-Tzer Yau: Taming Randomness 18

Life Sciences Simons Collaboration on the Origins of Life 20

Nurturing the Next Generation of Marine Microbial Ecologists 22 UNCERTAINTY BY DESIGN Global Brain Simons Collaboration on the Global Brain: Mapping Beyond Space 24 The theme of the Simons Foundation 2017 annual report SFARI The Hunt for Autism Genes is ‘uncertainty’: a concept nearly omnipresent in science 26 Simons Foundation Autism and mathematics, and in life. Embracing uncertainty, we Research Initiative SFARI Research Roundup designed the layouts of these articles using a design 28 algorithm (programmed with only a few constraints) that Sparking Autism Research randomly generated the initial layout of each page. 30

Tracking Twins You can view additional media related to these 32 stories by visiting the online version of the report Outreach and Education Science Sandbox: The Changing Face of Science Museums at simonsfoundation.org/report2017. 34

Math for America: Summer Think 36 COVER Leaders Scientific Leadership 38 This illustration is inspired by the interference pattern Simons Foundation Financials produced by the famous ‘double-slit experiment,’ which 42 provides a demonstration of the Heisenberg uncertainty Investigators principle. That principle, a hallmark of quantum physics, 44 states that there are fundamental limits to how much Supported Institutions scientists can know about the physical properties of a 59 particle. The more precisely scientists learn the momentum Advisory Boards of a particle, for instance, the less they can know about 60 the particle’s position, and vice versa. Board of Directors 63

Staff 64 LETTER FROM On September 6, 2017, in a small morning ceremony, the Simons Foundation inaugurated its new THE PRESIDENT in-house research division, the Flatiron Institute. The guest of honor, New York Gov. Andrew Cuomo, stirred the audience of scientists and mathematicians gathered in the institute’s lobby with the words, “I’m worried about the world that we’re leaving our kids. It’s much more complex, the problems are AND FROM much more difficult, and it’s only getting worse, and many of the solutions, I believe, are going to lie in the work that you do.” They were inspiring — and motivating — words indeed for a challenging THE CHAIR world full of uncertainty.

For tens of thousands of years human civilization has endeavored to reduce uncertainty and unpredictability. The development of agriculture addressed the question of where our next meal would come from, and the taming of fire assured us we would not freeze to death during an unexpectedly cold night. As time went on, the rain god and the witch doctor gave way to the meteorologist and the M.D. In recent centuries, science has played an increasingly large role in reducing uncertainty, and, in fact, to a great extent it is uncertainty itself that has driven science forward.

For the scientist, the concept of uncertainty includes the degree to which something is known. By systematically studying our natural and physical world, scientists chip away at uncertainty, replacing outmoded theories with more accurate ones. As Edwin Powell Hubble depicted this process and its concomitant mindset back in 1939, “The scientist explores the world of phenomena by successive approximations. He knows that his data are not precise and that his theories must always be tested. It is quite natural that he tends to develop healthy skepticism, suspended judgment, and disciplined imagination.”

Through our funding of basic science research, the Simons Foundation supports this rigorous process of investigation, this management of uncertainty. We are interested in fundamental questions about our universe, about life on our planet and about the mysteries of our own bodies and brains. And we are always intent on deepening our knowledge of mathematics, the lingua franca of science. By funding our own in-house computational research division, the Flatiron Institute, and by awarding grants to external scientists through their institutions, we strive to advance the frontiers of research in mathematics and the basic sciences.

In this 2017 annual report, you will read about the launch of the Flatiron Institute, and about the research being conducted there to better analyze neuronal activity, to predict the dynamical behavior of materials and molecules, and to analyze important astronomical events. From our grant-making division, you will read about how geoscientists are helping us to piece together the mystery of the origins of life. You will also see stories about the geolocation system of the brain, the basic science of autism, and how three Simons Investigators are grappling with the uncertainties inherent in physics and quantum computing.

The vision and guidance for all of these programs comes from an outstanding leadership team of scientific directors. In a special section, they each reflect on their work and the uncertainties they manage in helping this enterprise to do all that it does. Their deep scientific knowledge and managerial experience is invaluable to the foundation.

Every day, we feel eager to come to work at the foundation because it teems with intellectual activity — and we say that with certainty! We hope that as you read through this report you will enjoy the excitement of learning about the fruits of uncertainty.

Marilyn Hawrys Simons, Ph.D. Jim Simons, Ph.D. President Chair FLATIRON “The scientific output of The institute began as the Simons Center for Data Analysis (SCDA), launched in 2013 under Greengard’s leadership. Flatiron is remarkable and INSTITUTE SCDA focused on analyzing the increasingly vast datasets produced by biological research. With the success of this has exceeded my most venture, its scope expanded. In 2016, SCDA transformed optimistic expectations.” into the Center for Computational Biology and became the INAUGURAL first center of a new organization: the Flatiron Institute. That same year, the Center for Computational Astrophysics launched and began modeling the cosmos and analyzing CELEBRATION astronomical datasets. The youngest center, the CCQ, began operations in September 2017. Led by Georges and co- Within a brisk five-year period, the Simons Foundation’s director Andrew Millis, the center new Flatiron Institute has grown from the germ of an idea develops new numerical and floated at a foundation scientific retreat in 2012 to a large, analytical methods to solve the bustling hub for developing computational methods. In quantum many-body problem a ceremony on September 6, 2017, Simons Foundation and uses the solutions to predict co-founders Jim and Marilyn Simons and keynote work. Although several floors of state-of-the-art workspaces the behavior of materials and speaker New York Gov. Andrew Cuomo joined had already emerged from behind construction walls, molecules. The Flatiron Institute Flatiron Institute and Simons Foundation leaders to remodeling continued in the rest of the building. In the will name a fourth and final center dedicate the new research division of the Simons offices, researchers have already established the institute’s — focusing on a soon-to-be-decided Foundation. The event, held in the institute’s newly reputation with innovative software releases and important discipline — in 2018. renovated lobby, celebrated early progress and hopes scientific discoveries. for the future. A unique and crucial part of “I couldn’t be happier about progress in the Flatiron the institute’s framework is The Flatiron Institute is located just across the street from Institute,” Jim Simons says. “The scientific the Scientific Computing Core, the Simons Foundation and was launched to advance output of Flatiron is remarkable and has exceeded which develops the organization’s scientific research through computational methods, my most optimistic expectations.” computing infrastructure including data analysis, modeling and simulation. At full and collaborates with Flatiron capacity, the institute will house around 250 scientists The institute’s mission is an important one for New Institute scientists to create and

and programmers across four research centers and York, Cuomo said during the ceremony’s keynote address. implement new computational and Marilyn Simons, Gov. Andrew Cuomo and Jim Simons cut the ribbon during the Flatiron Institute’s inauguration a computing core. “This is a unique place where top Computational science “is where we need to grow and statistical tools for use across the ceremony. Also pictured, from left to right, are center leaders Leslie Greengard, David Spergel and Antoine Georges. people sharing a common interest in the growing impact flourish,” he said. “It’s not just what New York state scientific community. Co-directed of computation on science will exchange ideas across needs, it’s what the world needs.” by Nick Carriero and Ian Fisk, the core hosts an on-site Flatiron Institute researchers don’t have to apply for grants, freeing them scientific fields,” says Antoine Georges, director of the computing cluster comprising 7,000 cores. Two off-site to pursue long-term projects that might not be possible if continued new Center for Computational Quantum Physics (CCQ) The Flatiron Institute provides a permanent home for supercomputers at Brookhaven National Laboratory and funding were uncertain. The financial model also means that software at the Flatiron Institute. The institute “could lead to big professional scientists focused on the development of novel the San Diego Supercomputer Center at the University developed at the Flatiron Institute is freely available to all scientists and leaps forward in both fundamental understanding and new computational, mathematical and analytical techniques and of California, San Diego, lend additional processing is built to last, receiving long-term support and continued development. algorithmic methods.” software, says Leslie Greengard, director of the institute’s power, which will continue to grow. In aggregate, Flatiron That’s in contrast to other scientific software that typically is abandoned Center for Computational Biology. “Few existing academic researchers currently have access to 28,000 processor when funding runs dry or the student responsible for the code graduates. “We are already becoming a place where people come to institutions have developed a track for such people, cores hosted over the three computing facilities, a variety of learn new algorithms and approaches,” says David Spergel, especially at the scale of the Flatiron Institute. None have specialized resources — including general-purpose graphics The institute’s rapid growth has already shifted the staff composition of director of the Center for Computational Astrophysics at done so with such a broad scope in a single location.” processing units and visualization walls — and 10 petabytes the Simons Foundation as a whole, which, a short time ago, focused only the Flatiron Institute. “Because of the way we span fields, of storage space. on grant-making. The addition of an in-house research organization has we have the potential to be a unique place in transferring The Flatiron Institute’s New York home offers unique influenced other divisions of the foundation, notes Simons Foundation information, approaches and techniques between areas.” opportunities for collaboration, Spergel says. “It’s a place “The support of the Simons Foundation permits us, as best President Marilyn Simons. Institute scientists regularly mingle with where we’re surrounded by great universities and where we can, to ensure that the research community is only fellow foundation employees at lectures, lunches and staff meetings. The dedication event capped a year of hiring and startup we have the opportunity to partner with them.” In fact, limited by their imagination and their initiative, and not by milestones for the growing institution. Following the many senior researchers within the institute have joint the computing resource requirements and technology,” Fisk “Adding internal researchers to our organization has been inspiring to all ceremony, Cuomo toured the institute’s working areas and appointments at nearby universities, including New York said during the dedication ceremony. of us working at the foundation,” she says. “Their creativity, dynamism spoke with Flatiron Institute staff scientists about their University, Columbia University, Princeton University and intellectual depth has transformed our environment and given us all and Stony Brook University. “Our unfair advantage is the opportunity to share in the experience of scientific research.” New York,” he says. FLATIRON INSTITUTE FLATIRON

4 SIMONS FOUNDATION into the field through the Simons Collaboration on the Many Electron Problem, which Millis directs, and recognized the great need for new CENTER FOR computational methods to advance materials science.

COMPUTATIONAL Georges and Millis take a project-based approach to the organization of the CCQ. Some of the center’s current projects develop machine- QUANTUM learning methods, investigate strong light-matter coupling and build software tools. This structure gives researchers more flexibility as they Understanding and potentially controlling material navigate an uncertain path toward resolving the many electron problem, PHYSICS properties is difficult because of the inconceivable number says Angel Rubio, CCQ distinguished research scientist and managing of electrons involved. A quantum system comprises all director of the Max Planck Institute for the Structure and Dynamics of possible configurations, or states, of its particles, and the Matter in Hamburg, Germany. number of quantum states rises exponentially with the number of particles. A teaspoon of copper, for instance, “If you want to be revolutionary and open new avenues, you have to tackle contains more than 10 trillion electrons — that’s a lot things that are unknown,” he says. “Of course, you have an agenda; you of potential states. know what you want to solve. But in this process, you might deviate toward something that’s even more interesting than what you hoped The CCQ’s scientists can’t simplify the problem by to solve. That’s the beauty of CCQ — we’re not forcing the people to extrapolating from the behavior of a single electron. In keep the originally proposed path. We’re allowing the exploring of quantum physics, electrons can become entangled with one all possibilities.” another. Once entangled, particles can no longer be treated individually, even when physically separate. The collective One promising avenue for overcoming the many electron problem behavior of entangled electrons therefore drastically is the use of a mathematical concept called a tensor network. The complicates the problem. Currently, ‘brute-force’ methods method organizes information about a quantum system into bundles that incorporate entanglement by considering every called tensors. Quantum entanglement connects the tensors into a potential interaction between every electron can handle only network. For systems with a simple, regular structure, tensor networks two or three dozen electrons at once. Many useful material allow researchers to approximate the system using exponentially less properties, such as magnetism and superconductivity, only computational power. The approach is akin to the compression of video arise with millions or hundreds of millions of particles. files, says Stoudenmire, who studies tensor networks.

“There’s no way direct, brute-force simulation can work” for Shortly after the CCQ’s launch, Stoudenmire hosted a workshop that this problem, says Georges, who is also a professor at the attracted tensor network experts from around the world. The event, Collège de France. “It’s not a matter of waiting for better which included non-physicists, demonstrated that tensor networks offer computers. You’ll perhaps get an additional electron every unexpected potential for technologies such as artificial intelligence, he few years, so you’re not going to get to trillions of electrons says. “Physicists have been working in this niche area developing this anytime soon.” incredible mathematics just to solve this one tough physics problem that we have. But maybe along the way we’ve developed something more Luckily, alternatives to raw computational muscle exist. generally useful.” This illustration shows the lattice structure of anatase titanium dioxide along with a graphic representation (purple) of a 2-D exciton — an electron-hole pair — generated by the absorption of light. The Center for Computational Quantum Physics is investigating the new Under the leadership of Georges and center co-director quantum states obtained when many electron-hole pairs are generated and strongly coupled to light and other excitations of a material. Andrew Millis, also of Columbia University, the CCQ Fostering collaboration through events such as the tensor network Image courtesy of Joerg M. Harms of the Max Planck Institute for the Structure and Dynamics of Matter leverages computational methods to overcome the so- workshop is critical to the CCQ’s mission, Millis says. Events offer a called ‘many electron problem’ and predict the behavior of chance for CCQ members to mingle with other researchers in the field — From early stone tools to silicon computer chips, materials Overcoming the complexities of the quantum molecules and materials. The center’s focus on developing both theorists and experimentalists. have defined humankind’s progress. Electrons, discovered world presents a daunting challenge, which a new methods and algorithms sets it apart from research in 1897, are chiefly responsible for the physical properties new research hub aims to meet. The Center universities and government labs, Georges says. Those “This field needs close contact with experiment, or it becomes sterile,” Millis of molecules and materials. The behavior of these charged for Computational Quantum Physics (CCQ), institutions focus on shorter-term projects, whereas says. “It’s a testament to the poverty of our imaginations and the limits of particles determines why some metals hold an edge which launched at the Flatiron Institute in algorithm and code development can take years of continued our computational ability. Every few years, the field throws up a complete whereas others are pliable, why some substances react September 2017, is developing the theoretical tinkering and support. surprise, a class of materials that do something that nobody expected. With and others are inert, and why some materials conduct understanding, algorithms and computational better theory, we’re trying to reduce the number of those surprises.” electricity and others insulate. tools needed to bridge quantum mechanics and “There hasn’t been the right incentive structure in the field the behavior of molecules and materials. to keep the best programmers or algorithm developers If scientists could model and predict how large numbers of around,” says CCQ research scientist Miles Stoudenmire. electrons behave with one another, they could potentially “We are not a materials research center,” says “Programming just isn’t as important elsewhere. At CCQ, custom-design arrangements of atoms with fantastic CCQ director Antoine Georges. “That’s what it’s front and center.” properties, such as high-temperature superconductivity, makes us unique in comparison with places high-density energy storage and high-efficiency hydrogen that apply existing methods to materials. We’re The unfulfilled need for an organization such as the CCQ fuel generation. They might even uncover new properties taking the problem from its foundation, from helped drive its selection as the third Flatiron Institute that defy current understanding. the basic computational, mathematical and center. The Simons Foundation had already made inroads physical point of view. We’re devising new concepts and new methods.” FLATIRON INSTITUTE FLATIRON

6 SIMONS FOUNDATION CENTER FOR COMPUTATIONAL ASTROPHYSICS: NEUTRON STAR

MERGERS nucleus faster than the atom can undergo radioactive decay. organizer Matteo Cantiello, an associate research scientist Scientists believe at CCA. “We were able to rapidly attract people faster the r-process is Astrophysicists Eli Waxman of the Weizmann Institute in Israel and Jennifer Barnes of than other groups could, to talk about where we stand in responsible for much of the gold on Earth. If neutron star Columbia University talk during the neutron star merger meeting at the Flatiron Institute. terms of understanding this phenomenon that we’ve mergers are rare, the gold in a galaxy might be patchy, rather About 130 million light-years from Earth, the relics of two exploded stars never seen before.” than evenly distributed. neared the end of a spiraling, dyadic dance around each other. The dance partners were incredibly dense neutron stars: Just a teaspoonful of their The early fall meeting buzzed with excitement: The neutron Other observations agreed with theoretical predictions as neutron-rich star stuff has a mass of about 1 billion metric tons. star merger marked the beginning of the era of ‘multi- well, such as that neutron star mergers generate the bright messenger astronomy,’ in which both gravitational waves flashes of gamma rays that puzzled scientists for decades. “This is known to lead to correct scientific hypotheses,” joked one Over time, the stars drifted toward each other and picked up speed. Just and light can reveal insights into the same event. Theoretical astrophysicists, it seemed, had gotten a lot of of the participants. before collision, each orbit took fractions of a second. Then came the big things right. finale: a final merger that sent ripples through the fabric of space-time “This event brought us into a whole new regime of Some of the questions had a definite answer. Everyone agreed, for and the astrophysics community. On August 17, 2017, scientists at the understanding,” says Jennifer Barnes, an astrophysicist “The first day of the meeting, there was a feeling like there’s instance, that the event spotted in August was a merger of two neutron Advanced Laser Interferometer Gravitational-Wave Observatory (LIGO) at Columbia University. “For me, at least, this was the nothing to be done here. We can just pat our backs and head stars, rather than something else, like the merger of a neutron star in the United States detected the gravitational waves that emanated from first time I was exposed to some new ideas and home,” Cantiello says. That feeling didn’t last, though. “The and a black hole. that cosmic collision. interpretations of the event. The meeting highlighted day after, there were a lot of interesting talks that changed just how many open questions there are. It helped clarify the mood. There was a lot of controversy about how much Other questions, though, lacked clear-cut answers. The observed The event turned out to be the “cosmic gift that keeps on giving,” says the landscape of uncertainties and questions and we understand about aspects of this event.” gravitational waves generated by the merger reached peak intensity astrophysicist Samaya Nissanke, of Radboud University Nijmegen in the ongoing debates.” 1.7 seconds before the burst of gamma rays. Because both gravitational Netherlands. Observations of the merger ushered in a long list of firsts: The uncertainty built over the second day’s presentations, waves and gamma rays travel at the speed of light, the time gap The event was the first direct sighting of two neutron stars slamming In the immediate aftermath of LIGO’s detection of the culminating in an open discussion that evening. Brian was a surprise. together, the first confirmation that such collisions produce vast amounts neutron star merger, scientists paired the measurements Metzger, an astrophysicist at Columbia University, led the of heavy elements, such as gold, and the first time scientists identified with data from Advanced Virgo in Italy (another hunter of group through a list of significant questions about neutron “It’s an interesting question as to why we had such a delay,” says both gravitational waves and light coming from the same cosmic event. gravitational waves). The combined data allowed scientists star mergers, each accompanied by a list of possible answers. astrophysicist Bruno Giacomazzo of the University of Trento in Italy. The only previously detected gravitational waves had originated from to triangulate where in the sky the gravitational waves After each question, the scientists in attendance voted by a “It’s not something we expected, and it’s not something we can confidently merging black holes, which scientists don’t expect to produce light. originated. Within 11 hours of the initial gravitational wave show of hands on which answer they thought was correct. explain right now. We still have a lot of things we don’t understand.” observations, astronomers spotted the afterglow of the Those discoveries led 44 of the world’s top neutron star experts to gather neutron star merger in a galaxy about 130 million light-years That lack of certainty isn’t a bad thing, Cantiello says. “At the end of the in November at a three-day workshop hosted by the Flatiron Institute’s from Earth. “This was a stupendous event for astronomers,” meeting, people were happy because there seems to be an understanding Center for Computational Astrophysics (CCA) to discuss what scientists astrophysicist James Lattimer of Stony Brook University said both that scientists have done an amazing job and that there is more know — and don’t know — about neutron star mergers. at the Flatiron meeting. work to be done,” he says. “People want there to be stuff to be done. People have a passion for astrophysical puzzles. If everything is The meeting was the first time experts from around the world were able The first day of the meeting at the Flatiron Institute focused understood, you have to move on to another problem.” to assemble following the public announcement of the neutron star on the discovery and the scientific theories it quashed or merger a mere month before. Flatiron’s expertise, facilities and resources confirmed. Observations suggest, for instance, that the enabled the meeting to come together that quickly, says meeting neutron star merger spewed heavy elements such as silver, platinum and uranium into space, including the equivalent of 10 Earth masses’ worth of gold. That abundance of heavy elements points to a previously proposed mechanism called the r-process, in which neutrons cram into an atom’s In this artist’s illustration, as neutron stars collide (center), they fling material outward at nearly the speed of light in particle jets (pink). The fast-moving material generates a burst of gamma rays. Image courtesy of NASA’s Goddard Space Flight Center/CI Lab FLATIRON INSTITUTE FLATIRON

8 SIMONS FOUNDATION CENTER FOR

Pnevmatikakis soon brought on board COMPUTATIONAL Andrea Giovannucci, a neuroscientist who also has a Ph.D. in computer science. Giovannucci had been trying to interpret calcium imaging movies he’d made of mouse cerebellar granule cells, which BIOLOGY: are tiny and dense. “Recording them is like trying to distinguish specific voices in a full stadium,” Giovannucci says. With help from Pnevmatikakis, however, he was able to figure out how to adapt NEURONAL MOVIES Pnevmatikakis’ algorithm to this setting, making it possible, he says, “to single out sentences from the heavy background noise and overlapping voices.” Neurons found automatically by CaImAn (red) overlaid on an image representing active neurons in a mouse’s hippocampus as the mouse The pair realized that to make a widely useful tool, they would have to navigates a virtual reality environment. In late 2012, Eftychios Pnevmatikakis — then a postdoctoral look beyond Pnevmatikakis’ original tight focus on neuron detection. So Image courtesy of Eftychios Pnevmatikakis and Andrea Giovannucci; Data provided by Jeffrey Gauthier and David Tank of Princeton University researcher in statistical analysis of neural data at Columbia they have also examined how to correct for animal motion during filming, University — began working on a challenge he thought and developed tools to benchmark the software’s performance. would be key to neuroscience research in the coming years: how to handle the flood of data emerging from ‘calcium The resulting software, now called CaImAn, has been freely available to imaging’ of animals’ brains. the public since 2015 and has been widely adopted by neuroscientists who do calcium imaging. Meanwhile, Pnevmatikakis and Giovannucci The technology, which makes movies of animals’ brains by measuring light emitted from a protein as it binds to calcium ions in neurons, was starting to offer an exciting The pair realized that to make a widely useful tool, new window into living brains. The technique noninvasively captures the activity of large brain regions at single-neuron they would have to look beyond Pnevmatikakis’ resolution and allows researchers to control exactly which neurons, and even which type of neurons, are being original tight focus on neuron detection. measured. In recent years, the use of calcium imaging has exploded, with neuroscientists making movies of insect, are continuing to improve CaImAn. Most recently, in June 2017, they rodent and even primate brains. extended its functionality to identify neurons in real time, as data stream through the software frame by frame. The innovation means Yet Pnevmatikakis was concerned that the technology might that researchers can run CaImAn on ordinary computers that don’t have soon become a victim of its own success. Until a few years enough storage for the entire dataset. “We like to say that we develop ago, calcium imaging movies were small enough that a algorithms to do data analysis for the 99 percent,” Pnevmatikakis says. laboratory assistant could manually label the neurons, frame by frame. Today, however, it is common to collect 100,000

frames per hour, each containing thousands of neurons. An illustration of how CaImAn detects neuron activity. At left are source data frames of neuronal activity. CaImAn identifies the location of each discrete neuron (Component) within those frames as well as its signal, in red or blue, Pnevmatikakis developed an algorithm that could by differentiating the signal from background activity and noise, at right. automatically identify the neurons in certain datasets Image courtesy of Eftychios Pnevmatikakis and Andrea Giovannucci obtained from his colleagues at Columbia, and in 2014, he came to the Flatiron Institute (then called the Simons Center for Data Analysis, or SCDA) with a mandate to bridge the gap from his algorithm to a broadly useful software platform. “The mission statement of SCDA was to fill exactly this kind of gap, so I thought it would be a good match,” Pnevmatikakis says. FLATIRON INSTITUTE FLATIRON

10 SIMONS FOUNDATION CENTER FOR The human genome contains an estimated 19,000 genes. Those genes encode proteins that allow cells to carry out COMPUTATIONAL tasks such as ferrying oxygen molecules, fighting off diseases and communicating with fellow cells. But the HumanBase users can just type in a particular gene function of most genes remains elusive, and scientists are or disease and quickly receive a list of genes ripe for still struggling to crack the human body’s full genetic code. experimental scrutiny. For instance, if a gene often BIOLOGY: HUMANBASE expresses alongside genes already associated with In March 2017, researchers at the Flatiron Institute’s increased risk of Parkinson’s disease, that gene could Center for Computational Biology (CCB) soft-launched a be a tempting target for further research. “It’s guilt new tool for decoding the human genome. The cloud-based by association,” says project leader and CCB data software, called HumanBase, uses machine learning to trawl scientist Aaron Wong. through decades of genomic research data for previously unseen potential biological connections. HumanBase can The algorithms that power HumanBase have already sleuth out how specific genes could potentially control cell proved their prowess. In August 2016, Troyanskaya, functions, influence the expression of other genes, and Wong and their colleagues reported online in contribute to disorders such as autism. Researchers can then Nature Neuroscience a substantial breakthrough in carry out experiments that verify those potential connections. the hunt for genes associated with autism. Using a predecessor to the numerical tools employed in “There’s a huge wealth of undiscovered knowledge in these HumanBase, the researchers identified roughly data,” says Olga Troyanskaya, deputy director for genomics 2,500 genes potentially linked to autism. Several of at CCB. “We wanted to build a single resource that could the most promising candidate genes had no prior help biologists discover and leverage that knowledge.” genetic research tying them to autism. Scientists had previously identified 65 autism risk genes and Historically, knowledge in the field of genomics largely predicted that 400 to 1,000 genes are likely involved rested in published findings and the heads of biologists. in autism susceptibility. That’s changed, Troyanskaya says. New experiments now generate colossal datasets, “A lot of these connections you just cannot but genetic associations are often too faint to uncover with see with traditional approaches.” any certainty from any one experimental dataset. Finding those connections requires HumanBase users are already tapping the software’s looking at a much bigger picture, Troyanskaya says. potential to generate new hypotheses and spark new experiments, but the software’s development is far “A lot of these connections you just cannot see with from over. Troyanskaya, Wong and their team plan to traditional approaches,” she says. “You need computational add even more datasets to HumanBase’s knowledge algorithms that can pick up granules of data across multiple bank every six months and to continue developing datasets. That’s impossible to do just in your head.” the algorithms powering the software. “We want to build something that biomedical scientists can HumanBase incorporates data from more than 38,000 rely on,” Wong says. “We want them to incorporate genomic experiments and more than 14,000 scientific HumanBase into their research workflow — publications. The software standardizes all of those data to drive new hypotheses and follow-up experiments.” before trained algorithms sift through the information looking for biological connections, particularly in the context of specific tissues, cell types and diseases.

An analysis generated by HumanBase displays genes functionally related to KMT2A, a gene associated with autism. Such analyses can help scientists uncover previously unseen connections between genes and disorders. FLATIRON INSTITUTE FLATIRON

12 SIMONS FOUNDATION A diagram illustrating the relationship of several complexity classes in theoretical computer science. The BQP (bounded-error, quantum, polynomial time) class consists of problems that quantum computers can solve in polynomial time. The BQP class SCOTT AARONSON: may extend outside of NP, meaning that a quantum computer could solve certain problems faster than traditional computers could even verify the answer. Despite the promise of quantum computers, some problems, such as those in the NP-complete QUANTUM AND class, remain unsolvable in polynomial time. Illustration adapted from images by MIT CLASSICAL OpenCourseWare, Scientific American UNCERTAINTY

Or so it would seem. But at the heart of any discussion about the relative merits of quantum versus classical computing is a nagging uncertainty called ‘P versus NP.’ This problem For a quantum computing optimist, concerns how quickly the amount of time it takes to solve Scott Aaronson spends an impressive amount a problem grows as a function of the input size. There are of time trying to figure out what quantum some problems — factoring large numbers is one famous computers can’t do. A computer science example — that no one knows how to solve in what is known The fact that P versus NP and related questions are still professor at the University of Texas at Austin as polynomial time. (That is, the time to run the algorithm unsettled makes the true extent of quantum advantages and a Simons Investigator, Aaronson pushes is a polynomial function of the size of the input, such as difficult to ascertain. In the 1990s, Peter Shor developed an the boundaries of both classical and quantum n2 or n500,000.) Instead of polynomial time, the best-known algorithm for factoring numbers in polynomial time using quantum computer will move more quickly. Computations that take computing to better understand what they can classical algorithms to solve these problems have a number a quantum computer. It is leaps and bounds more efficient seconds to perform on a 50-qubit computer may take days to verify on a and can’t do. “I try to understand the ultimate of steps that grows exponentially with the size of the inputs. than the best-known classical algorithms, but that doesn’t classical computer. But verifying the result of computations on a 100-or limitations of algorithms,” he says. “It goes hand For large inputs, exponential growth completely overwhelms necessarily mean there aren’t similarly efficient classical 200-qubit system could very well become functionally impossible for a in hand with understanding what is possible.” polynomial growth. algorithms no one has yet been clever enough to find. classical computer.

Classical computers run on bits. Each bit can be Problems that are known to have polynomial-time solutions Shor’s algorithm for factoring large numbers so far seems When people think about the applications of quantum computing, they 0 or 1, on or off. Quantum computers instead are considered ‘easier’ than problems that don’t have to be a special case. For most problems, Aaronson and other often start worrying about the security of current cryptographic systems. use quantum bits, or qubits, that can exist in such solutions, which computer scientists colloquially call researchers have found that quantum computers have only For example, RSA (Rivest–Shamir–Adleman), a common public-key superpositions of states. In a superposition ‘hard.’ But for some of these hard problems, including a limited, though often substantial, speedup over classical cryptosystem, relies on the difficulty of factoring large numbers for of states, each state has a number called an integer factoring, it is easy to check whether a proposed algorithms. So far, other than for factoring, quantum its security. Aaronson points out that, although eventually quantum amplitude associated with it. In some ways, an solution is correct. The best-known algorithms for integer simulation and a few other tasks, quantum algorithms computers could break existing public-key cryptography, there are other amplitude is similar to a probability, but instead factoring take exponential time to implement, but it only usually cannot take an exponential-time problem and turn cryptographic systems waiting in the wings that we don’t know how to of the positive probabilities we are familiar takes polynomial time to multiply two numbers and check it into a polynomial-time problem. break — not even with a quantum computer. with, these numbers are complex. (Complex whether their product is the desired number. “We don’t have numbers consist of numbers of the form x + iy, the mathematical tools right now to prove unconditionally Google, IBM and other research teams have recently built Aaronson is even more excited about another application of quantum where x and y are both real numbers and i is that most of the problems we think are hard really are hard,” 50-qubit quantum computers, with about twice as many computing: simulation. What better to simulate complicated chemical the square root of -1. Complex numbers provide Aaronson says. The P versus NP problem, which has a qubits and many orders of magnitude more computational interactions, down to the quantum mechanical level, than a computer a two-dimensional rather than one-dimensional million-dollar bounty on its head, asks whether the problems power than what had previously been available. (Data about that runs on quantum mechanics itself? “Once you’re confident the set of numbers to work with.) By carefully that seem hard to solve but easy to check are actually hard the performance of these computers may not be available device is working, you can start using it to simulate molecules that manipulating the amplitudes associated with to solve at all. for some time.) Past the 50-qubit point, another uncertainty have never existed,” he says. While some of the problems that quantum different states so that correct answers are starts to creep in. If you start using quantum computers to computers could solve seem to be hard for the sake of being hard amplified and incorrect ones cancel out, a Aaronson and his collaborator Avi Wigderson have also done solve computationally expensive problems that aren’t easy (indeed, much cryptography relies on such questions), simulating quantum algorithm can achieve a massive substantial research on what is known as the algebrization to check, how do you check your work? “There’s a bit of complicated chemical interactions could have a tangible impact on a speed advantage in calculations for some barrier, which they identify as one of the main obstacles to irony here,” Aaronson says. “For the problems that we know range of questions in physics, chemistry and biology. Although quantum- problems compared with a classical computer. showing that P and NP are indeed different. how to attack in the near future, the only way we know how computing-aided drug discoveries are still a long way off, the next few to verify the quantum computer’s results is essentially to years could lay the groundwork for the application of quantum physics to simulate the whole computation on a classical computer. several areas of scientific research. But the very fact that you can do the latter calculation means that the quantum computer isn’t doing something that’s impossible for classical computers.” Of course, the MATHEMATICS AND PHYSICAL SCIENCES MATHEMATICS

14 SIMONS FOUNDATION Since that discovery, Glotzer’s group has been studying different particle shapes to create different crystal and quasicrystal structures using only SHARON entropy. “We’re marching through the database of crystal structures. We’re finding another crystal structure, and then another one, and then another one,” she says. They would like to determine whether there are any crystal structures that can’t arise solely through entropy. “Just how GLOTZER: powerful is entropy as a driving force for order?” Glotzer asks. In addition to the intellectual curiosity it conjures, this research could someday help people engineer materials with desired properties, such as photonic crystals: materials that trap certain wavelengths of light used in fiber-optic cables. To create these new materials, researchers will ORDER FROM need to harness other forces aside from entropy, but Glotzer’s research helps explain what role entropy has in the process. “If I want to use only entropy, what shape do the particles have to be?” she says. “Once we understand that, we can combine entropy with other forces to get UNCERTAINTY precisely the material we want.” “I’m trying to argue for elevating the recognition and respectability of the entropic bond.” In 2009, Glotzer’s research group made a similar discovery for tetrahedrons rather than spheres. While working with Another goal is to quantify — and spread the idea of — the entropic The second law of thermodynamics says that in a closed system, entropy semiconducting nanoparticles that happened to have a bond. Glotzer’s work shows that entropy can be thought of as similar — colloquially thought of as a measure of disorder or uncertainty tetrahedral structure, they found that when the packing to chemical bonds between atoms, such as hydrogen bonds or metallic — cannot decrease. Generally, higher entropy corresponds to higher fraction got high enough, a quasicrystal structure emerged bonds. Those bonds are based on electron density; entropic bonds temperatures, phase transitions from solid to liquid or liquid to gas, with 12-fold rotational symmetry. “It was a completely are based on entropy density. “I’m trying to argue for elevating the or a general decrease in the order of molecules in a system. Simons serendipitous discovery that we were never looking for, recognition and respectability of the entropic bond,” she says. “When Investigator Sharon Glotzer, professor of chemical engineering at the nor did we expect,” Glotzer says. “We didn’t believe it for you go to ‘bond’ in Wikipedia, it should list all these bonds, and one of University of Michigan, studies the counterintuitive instances when a long time, until we proved to ourselves it was real.” them should be ‘entropic bond.’” entropy doesn’t work that way. “I’m interested in how entropy can lead to order rather than disorder,” she says. “The opposite of what most Once again, order had emerged solely as a result of entropy, people think about when they’re thinking of entropy.” not from molecular interactions. But this time, the team ended up with a quasicrystal instead of a crystal structure. Entropy appears in different scientific disciplines and in various guises, Unlike crystals, which have periodic structures that repeat but Glotzer is concerned with Gibbs, or statistical, entropy. In essence, precisely, quasicrystals have an ordered structure but this is a measure of the number of distinct arrangements, also called no exact translational symmetry and are generally more microstates, possible in a system. “Typically, the states that we observe complicated than crystals. “It was amazing, because here in nature are the ones for which there are the most possible we were looking at one of the simplest three-dimensional arrangements,” Glotzer says. solids, the tetrahedron, forming one of the most complicated structures possible — a quasicrystal!” Glotzer says. To illustrate: If you put a collection of identical hard spheres — marbles or tiny nanoparticles — in a large container and send it into space to remove the effect of gravity, the spheres will tend to spread out randomly. But if you reduce the size of the container, you’ll start to see strange results. When the proportion of the container occupied by the spheres increases to 50 percent, the spheres will form a regular lattice, or crystal structure, even though there is room for them to remain in a less ordered arrangement. “The idea is that when particles are sufficiently crowded, there are more patterns that are crystalline than patterns that are disordered,” Glotzer says. Counterintuitively, the particle arrangement with the most entropy is highly ordered. In essence, entropy has driven the emergence of order.

Hard particles can arrange themselves into crystal structures through entropy alone, Sharon Glotzer and colleagues discovered. Illustration adapted from images by P.F. Damasceno et al./Science 2012 MATHEMATICS AND PHYSICAL SCIENCES MATHEMATICS

16 SIMONS FOUNDATION His boldest dreams, though, are about finding applications of random matrix theory beyond HORNG-TZER the quantum world where Wigner’s work started. With computing power and dataset sizes at all-time highs, the networks available the values were distributed normally, but for analysis are more massive than ever. Yau other distributions remained elusive. Wigner hopes that the techniques he has developed in YAU: TAMING and other physicists who studied the question random matrix theory will offer insight into the believed that the behavior of these large inner workings of neural networks — computer random matrices would be the same no systems that ‘learn’ by finding patterns in matter the distribution from which the entries data. So far, even finding an appropriate were drawn. mathematical language to describe what a RANDOMNESS neural network does has eluded mathematics. On its face, that was a startling claim. How “Mathematicians are not even sure what the could properties based on a particular random right question is yet,” he says. studying random matrices. For many years, he worked distribution turn out to be independent of that on a variety of problems related to statistical mechanics, distribution? But there was a precedent for On the other hand, neural networks are large, The matrix is a workhorse of modern math — and of including fluid dynamics and complicated many-body such a situation: The central limit theorem complicated correlated systems — exactly the physics, computer science and engineering, for that matter. systems that arise in astrophysics. In 2000, he was awarded in statistics asserts an analogous result. In type of systems Wigner considered. Although This mathematical object is an array of numbers that a MacArthur fellowship, popularly known as a ‘genius grant,’ essence, Yau and colleagues showed that the precise mathematics of how neural encodes a transformation between mathematical spaces. in recognition of the contributions he had made to both Wigner’s conjecture was a sort of central limit networks are related to random matrices is Matrices can represent the way various elements of a system mathematics and physics in studying those problems. theorem for random matrices. unclear, Yau is hopeful that new insights from rely on one another. Among researchers who study social the study of random matrix theory will allow networks, for example, an adjacency matrix that represents Shortly before starting to tackle Wigner’s conjecture, These days, Yau is extending his work in several his team — or other researchers — to tackle the the connections between people will use a 1 to indicate Yau had been focusing on questions related to the directions. He hopes to circle back to the modern challenges in understanding neural that two people in the network know each other and a quantum mechanical behavior of semiconductors using work on Anderson’s model of semiconductors, networks and machine learning. 0 to indicate they are strangers. random Schrödinger operators. This work built on that eventually determining whether they really of Nobel Prize winner Philip Anderson, who modeled are governed, as Anderson predicted, by the In the 1950s, physicist Eugene Wigner began to use semiconductors with impurities as lattices with randomly same rules as random matrices. Yau sees the matrices to model the interactions of atomic nuclei. distributed obstacles. As Yau was working on Anderson’s semiconductor as an important example to These interactions are so complicated, particularly in a model, he found that he kept hearing that Wigner’s random study, in part because the randomness of the system large enough to represent anything of real-world matrices were lurking in the background. “The question was system is entirely natural rather than imposed “You sometimes significance, that it is impossible to discern the exact always mentioned in meetings,” he says. Eventually, he and by humans. “Most random matrix examples are matrices involved. Instead, he had the insight that random some of his co-investigators decided to alter course and work constructed by something like flipping a coin, a stumble on things in matrices could be used instead. The idea of a random matrix on Wigner matrices themselves. human construction,” Yau says. “The Anderson is that instead of looking at a matrix with certain fixed values, model is so precisely given by a law of nature.” unexpected ways.” one imagines a matrix in which all values are selected In order to say anything meaningful about a matrix full of randomly from a set of numbers with fixed parameters. random numbers — indeed, to say anything meaningful Although some of Wigner’s beliefs about Wigner believed these random matrices would behave about random numbers at all — the researcher must make random matrices have been confirmed, there in particular ways, shedding light on the systems they assumptions about how the numbers are chosen. Are the are as yet no examples of physical systems were modeling. numbers distributed that conform to those random matrices. normally — the Yau hopes to construct a quantum mechanical A few years ago, Simons classic bell curve model that will actually display the behaviors Investigator and — or taken from of Wigner’s random matrices. Harvard mathematics some more exotic professor Horng-Tzer distribution? Not Whereas Wigner’s original vision for random Yau and his collaborators long after Wigner matrices came from his work on quantum successfully answered posed the problem, mechanics at the level of atomic nuclei, key questions Wigner Freeman Dyson and Yau’s work has taken that vision on a winding had posed, closing that other physicists were path away from quantum mechanics into pure chapter in the book of able to successfully mathematics. It may well continue into the random matrices. Yau probe the case where decidedly macroscopic realms of human social didn’t start his career networks and other forms of data analysis. “That’s how it happens in science,” Yau says. “You sometimes stumble on things in unexpected ways.” Histogram of eigenvalues of a real symmetric matrix of large size with random entries. The distribution follows Wigner’s semicircle law. MATHEMATICS AND PHYSICAL SCIENCES MATHEMATICS

18 SIMONS FOUNDATION In 2017, SCOL invited several earth and planetary scientists to give presentations at collaboration meetings for consideration to become investigators, signaling the collaboration’s appreciation of the necessity of geoscience in work on the origins of life. Insights provided by the SIMONS COLLABORATION collaboration’s chemists and biochemists, such as a potential What rocks on ancient Earth looked like remains somewhat uncertain, need for ultraviolet light, can help guide the geoscientists though. Plate tectonics, weathering, erosion and biological activity have ON THE toward environmental clues. For instance, exploring the destroyed nearly all relics of early Earth’s surface. Luckily, scientists do processes and molecules that would have been available know of a treasure trove of remnant rocks — it’s just not on Earth. in streams and pools on early Earth could help chemists ORIGINS OF LIFE further hone their experiments. “What I like about this Mars, unlike Earth, has undergone relatively little resurfacing. The is that the chemists on their own couldn’t do it, the Curiosity rover discovered that rocks from the planet’s Gale Crater and geochemists on their own couldn’t do it,” Sutherland says. surrounding landscape date back between 3.6 billion and 4.1 billion years “We need to work together.” — making them around the same age as the oldest rocks on Earth. “We can look at sedimentary rocks on Mars that are effectively the same age The search for life’s origins poses a challenge, as it Bridging the gap between pristine laboratory experiments as the very oldest sedimentary rocks on Earth that may contain life,” says incorporates aspects of many different fields. The breadth of and real-world conditions requires making the experiments SCOL investigator John Grotzinger, who will also become a SCOL uncertainties involved, from the composition of the ancient “dirtier,” says SCOL investigator Dieter Braun, professor of co-director in 2018. He is a former project scientist for the Curiosity Roughly 4 billion years ago, the early Earth was Earth’s atmosphere to the mechanisms that assembled the biophysics at the Ludwig Maximilian University of Munich. rover mission and Fletcher Jones Professor of Geology at the California an unfamiliar world. Large-impact craters pitted first biomolecules, places the subject out of reach for any Inspired by work done by geologists, his group experiments Institute of Technology. the planet’s surface, the sun glowed dimly, and one research group or lab to fully explore, Sasselov says. using 3-D-printed rock replicas in the lab. Braun and his the atmosphere was almost entirely devoid of colleagues mix powdered minerals such as basalt with In June 2017, Grotzinger and his colleagues reported in Science that oxygen. Yet from this extreme environment, the The need for a large, collaborative hunt for the origins of printing resin. Using a 3-D printer, they can layer custom geochemical evidence from Curiosity’s trek through the Gale Crater oldest known evidence of life appeared: Mound- life sparked interest during a 2012 Simons Foundation shapes of the mixture onto polished rock. The artificial suggests the region once hosted liquid-water lakes with layering similar like structures built by microbial communities, meeting at the Buttermilk Falls Inn in Milton, New York. rocks mimic the porous volcanic rocks that were probably to those found on Earth. Sediments near the top of these ancient lakes crystals containing isotopic traces of biological That gathering assembled experts from many disciplines abundant on the early Earth. The powdered rocks can contain evidence of more abundant oxygen than what was found in those activity, and microfossils embedded in ancient to discuss potential Simons collaborations. The next catalyze and take part in chemical reactions on the surface, closer to the lake bed. This division offers further indication that such rocks all hint at when the first earthlings year, SCOL began with 12 founding investigators and one and small pores in the mimic rock can trap water, create lakes may have been favorable to life. emerged. But how Earth went from lifeless honorary investigator. Over the next several years, the air pockets and stash molecules that would otherwise get to lush remains uncertain. collaboration added more members to bolster its collective washed away. “If you find a lake on Mars and you compare it to an ocean on Earth, it’s expertise and pursue new research paths. As of 2017, the not exactly the same,” Grotzinger says. “It’s certainly not an identical twin, “It’s the biggest question that science has group had 24 investigators and 14 postdoctoral fellows. “If we just keep our experiments in glass flasks and clean but it’s probably a cousin.” The Martian lakes can offer insights into what not been able to come to grips with,” says “We started as a bunch of individuals working on the things environments, we will never find these solutions,” Braun similar bodies of water may have looked like on the early Earth, he says. Roger Summons, professor of geobiology at that interested us, the things we’re capable of doing on our says. “I think we have to go into those dirty geological Dissimilarities between the two can shed light on why the two planets the Massachusetts Institute of Technology. own,” says Summons, a SCOL researcher and member of settings, and I think we’ll learn a lot by doing that.” ultimately diverged in terms of habitability. “It’s a grand challenge.” the collaboration’s steering committee. “But as we got to know each other and talk and meet and cogitate, people That challenge has attracted an international came up with ideas that otherwise would have never group of scientists, each bringing expertise happened without this collaboration.” “We’re looking for any plausible pathways, not from disparate fields such as molecular biology, exoplanetary science and geochemistry. The The collaboration is pursuing many potential avenues for necessarily for the one and only origin of life here Simons Collaboration on the Origins of Life the origins of life, and one thread is generating particular (SCOL), which started in 2013, explores how interest. In a 2015 paper in Nature Chemistry, SCOL on planet Earth.” life can arise and evolve, whether on Earth or investigator John Sutherland of the MRC Laboratory elsewhere in the cosmos. of Molecular Biology in the United Kingdom and his SCOL’s ongoing research into all aspects of colleagues demonstrated the creation of precursors to the emergence of life, from chemical bonds to “We’re talking about origins — plural — of nucleic acids, lipids and amino acids — basic building extraterrestrial lakes, makes Sasselov optimistic life,” says Dimitar Sasselov of Harvard blocks of all known life-forms — from hydrogen cyanide, that the field is closing in on an origin story. University, who co-directs the collaboration hydrogen sulfide and a dash of ultraviolet light. “We are terribly close, it seems to me, to having with Jack Szostak of Harvard. “We’re looking the whole thing finished for one pathway,” he for any plausible pathways, not necessarily The necessity of ultraviolet light limits where these reactions says. “But the jury is still out. It may turn out for the one and only origin of life here on could have occurred on early Earth, says Sutherland, who to completely be a dead end and back to the planet Earth.” will become a SCOL co-director in May 2018. Sunnier spots, drawing board, but it doesn’t seem like that to such as streams and pools on land, would better facilitate me. So many things have fallen into place.” such reactions than dimly lit locations such as hydrothermal vents in the deep ocean. “And at that point, you need to start talking to the geochemists,” Sutherland says.

Biophysicist Dieter Braun presents during the

LIFE SCIENCES Simons Collaboration on the Origins of Life annual meeting.

20 SIMONS FOUNDATION “If your training is in a NURTURING THE physical field, it can be hard NEXT GENERATION to move into ecology. We Some of the 2017 awardees have doctoral degrees related to thought providing funding oceanography. But the fellowship also encourages applicants OF MARINE from quantitative areas, such as physics and mathematical would encourage people to modeling, because the interdisciplinary nature of marine make that transition.” microbial ecology means that quantitative research can MICROBIAL ECOLOGISTS have a significant impact. “If your training is in a physical field, it can be hard to move into ecology,” Carlson says. “We thought providing funding would encourage people to make that transition.” Natalie Cohen is using her Simons fellowship at Woods Hole Oceanographic Institution in Massachusetts to Keisuke Inomura, a Simons Fellow at the University of study how marine microbes adapt to changing supplies Washington who is exploring mathematical models of the of trace metals. The other postdoctoral opportunities she distribution of different nitrogen-fixing microbes in the had applied for involved joining existing research projects ocean, says the fellowship offers him “more possibilities instead of designing experiments based on her own specific and more degrees of freedom” than he otherwise would have. interests, she says. Many postdoctoral fellowships last only one year, and even Oceans cover more than 70 percent of the planet. And within in a two-year position, “you have to get results very quickly “The day I got my acceptance letter for the Simons the oceans, it is the microbial ecosystem, made up of bacteria — like, in one year — if you want to be a strong candidate for fellowship was the best day of my life, because it meant I and phytoplankton, that forms the base of the food chain and an assistant professor job,” Inomura says. Having a three- could continue to study what I was most passionate about,” powers the cycling of carbon, oxygen and other elements. Yet year fellowship essentially doubles that time frame, he says, she says, and three years’ worth of funding “means that despite its importance, marine microbial ecology remains an lifting the pressure and allowing fellows to pursue more instead of spending my time writing grants, I can put all underfunded area of science. ambitious projects. my efforts into doing the best science I can.”

To help remedy this situation, in 2017 the Simons Foundation created the Simons Postdoctoral Fellowships in Marine Microbial Ecology. These three-year awards, given out each year to five early-career researchers, were inspired by Sallie (Penny) Chisholm, a marine microbiologist at the Massachusetts Institute of Technology (MIT). “She always gets lists of fellowships that MIT postdocs can apply to, and there’s never anything targeted to people in this field,” says Marian Carlson, director of the Simons Foundation’s Life Sciences division.

Keisuke Inomura modeled how material and “For this area to thrive, we have to get young scientists to find energy flow through Crocosphaera, single- it an area where their careers can flourish,” Carlson says. celled marine cyanobacteria. The microbe (circle) accumulates carbon during the day. At night, it uses that carbon to expel oxygen. In choosing fellows, the foundation pays particular attention The resulting low oxygen levels within the to whether the applicant’s research plan and proposed mentor organism permit the conversion of nitrogen into ammonia, an important part of the will help the applicant transition to doing independent marine nitrogen cycle that has rarely been research of the highest quality. “This is a career award, not mechanistically quantified. Illustration adapted from images by a grant to do a specific project,” Carlson says. “We wanted Keisuke Inomura

Natalie Cohen prepares tubing to pump seawater with to choose candidates who would move the field forward for minimal trace-metal contamination from the northeast the next decade.” Pacific Ocean to examine how phytoplankton respond to rapid changes in available iron. Image courtesy of Benjamin Twining of Bigelow Laboratory for Ocean Sciences LIFE SCIENCES

22 SIMONS FOUNDATION The researchers discovered a set of cells that act very much like place cells. Instead of firing when the animal is in a specific location, these determine which stimuli best predict whether ‘sound cells’ fire when the animal hears a specific tone. The findings a neuron will fire. “We’re trying to break free were published in Nature in March 2017. of a preconceived notion of how the brain is operating,” Giocomo says. “With our approach, Tank, Aronov and others have also found that neurons in the you can take a blind perspective.” hippocampus seem to respond not just to space or sound, but to every aspect of the task, firing in a predictable sequence throughout an With this system, Giocomo’s team has been experiment. “From pressing the lever to traversing the frequency to able to identify what 75 to 90 percent of cells receiving the reward and starting a new trial — there is a sequence of in the region respond to, compared with less activation throughout the entire period of behavior,” Tank says. than 50 percent using traditional methods. “We are now capturing a huge percentage of Taken together, the research suggests that cells in the hippocampus SIMONS COLLABORATION what the entorhinal cortex is doing during and entorhinal cortex are far more flexible than scientists thought. “The behavior,” Giocomo says. findings point to a more general function of the hippocampus,” says ON THE GLOBAL BRAIN: Aronov. Rather than simply encoding where an animal is or how fast it’s The analysis, published in Neuron in April moving, the neurons in this brain region map out whatever variables 2017, revealed that many cells in the entorhinal seem to be most important in that context. “If the animal is navigating MAPPING BEYOND SPACE cortex are both complex and flexible. A cell through space, the sequence [of electrical activity in this group of might code for one property when an animal neurons] tends to correspond to space. If it’s traveling through sound, is running slowly and another when it runs the sequence will correspond to successive sounds.” fast. “That’s a fundamentally different way of Neuroscientists have traditionally defined cells thinking about how the brain might compute Researchers theorize that this flexibility, also known as remapping, helps in the hippocampus and entorhinal cortex location in space,” Giocomo says. “What the animals encode experiences much more broadly. Remapping helps according to the specific navigational ability field has defined as grid cells is probably on one the brain “distinguish between experiences with strongly overlapping they seem to support. Neurons known as grid end of [the] spectrum — it’s just the tip of the elements, such as parking your car in the same parking lot but in cells track general location in space. Place cells iceberg in terms of coding features.” different parking spots each day,” Giocomo says. encode specific locations, firing whenever you pass your house, for example. Still others act Dmitriy Aronov of Columbia University Giocomo and Tank are now working together, along with Surya Ganguli, Whether you’re lost in a new city or driving the as speedometers, encoding how fast you are and David Tank have also found striking Loren Frank, Elizabeth Buffalo, Uri Eden and Ila Fiete, on a new SCGB most familiar roads, both your hippocampus moving. “People envisioned the system as a flexibility in both the entorhinal cortex and project that will delve more deeply into the remapping process. Their and your entorhinal cortex are hard at work. GPS: The brain knows the speed and direction the hippocampus. The team showed in 2014 project will expand on previous research, which focused on how neurons Together, these two brain regions create a you’re traveling from specific cells and can then that animals learning to navigate a virtual remap in response to visual changes in the environment — if the wall powerful human navigation system, with determine distance,” Giocomo says. reality world shifted their internal maps when color in a room changes or a wall is knocked down, for example. Giocomo diverse cells coordinating to perform different experimenters shifted that virtual world. They and Tank’s work shows that remapping can happen quickly, such as navigational functions, such as tracking reasoned that if animals can readily adapt to a when an animal changes its behavior, and across different modalities, the location, speed and direction of your “We’re trying to break free of a new spatial reality, maybe they can map other such as sound. “We want to use that as a springboard to understand how movement. Discovering this navigational types of experiences as well. neurons remap, what time course this remapping follows and how this system was a huge feat in neuroscience, information then gets communicated to the rest of the brain to form a earning John O’Keefe, May-Britt Moser and preconceived notion of how To test that theory, Aronov, then a postdoctoral memory or drive behavior,” Giocomo says. Edvard Moser a Nobel Prize in 2014. researcher in Tank’s lab, trained rats to traverse the brain is operating.” an auditory rather than a physical space. The Giocomo, Tank and their collaborators plan to record electrical activity But a growing body of research suggests these That approach has drawbacks, however. It animals used a joystick to move through a from the same neurons as animals perform two different tasks — brain regions may play an even more expansive requires researchers to hypothesize a priori as sort of soundscape — a defined sequence of foraging for food in an open environment and finding their way through role in how the brain organizes experience. to what they believe a cell is doing — such as frequencies. Moving the joystick to the left, for a maze — and analyze how neuronal activity differs depending on New findings from David Tank of Princeton encoding speed or location — and look for cells example, might increase the frequency. the environment and the task. They aim to uncover how remapping University, who is director of the Simons that respond to these variables. To find place helps the brain encode unique experiences and generalize across Collaboration on the Global Brain (SCGB), and cells, for example, scientists look for cells that different experiences. others show that the cells of the hippocampus fire most when the animal is in a particular and entorhinal cortex can encode much more location but grow quiet as the animal moves than physical space: They can also track away. The problem is that most cells in these sound and other factors. And Lisa Giocomo, a regions don’t behave in such a predictable neuroscientist at Stanford University and an way and therefore cannot be labeled SCGB investigator, and her collaborators have with a function. As a rat runs along a straight track, different neurons in the rat’s hippocampal region fire. shown that these cells seem to be capable of (As the rat enters the green portion of the rapidly changing their coding properties to Giocomo and her collaborators developed an track, for instance, the green neurons fire.) Researchers later conducted a similar respond to a novel environment or task. assumption-free method to define cells’ roles. experiment in which rats manipulated a Rather than looking for cells that code for joystick to change the frequency of a tone. The work showed that neurons respond to predefined spatial properties, the researchers changes in sound frequency in a manner have developed statistical models that similar to their response to changes in location. Illustration adapted from J.W. Rueckemann and E.A. Buffalo/Nature 2017 GLOBAL BRAIN

24 SIMONS FOUNDATION

Researchers supported by the Simons Foundation are also taking a closer look at ‘synonymous’ mutations among individuals in the SSC — mutations that, by definition, don’t One approach is simply to sequence more families. change the sequence of amino acids in the coded protein. ‘Whole-exome’ sequencing — sequencing the protein-coding It might seem, at first glance, that these mutations should regions of the genome — remains an effective way to not confer heightened autism risk, because they don’t identify new autism risk genes and strengthen evidence change which protein gets created. Yet these mutations THE HUNT for existing candidate genes, says Louis Reichardt, do sometimes affect either the ‘splicing’ or the amount SFARI’s director. “It works, and there’s a lot more to be of turnover of the messenger RNA molecules that carry discovered there.” the gene’s instructions to the ribosomes, the cell’s protein factories. These changes can in turn affect the amount of To date, researchers supported by SFARI and other protein that gets created. “It’s important to try to find as FOR AUTISM institutions have carried out whole-exome sequencing on many of these mutations as we can,” says Alan Packer, a about 6,000 families. A SFARI initiative called SPARK, senior scientist at SFARI. or Simons Foundation Powering Autism Research for Knowledge (see page 32), aims to raise that number to Searching for somatic mutations 50,000. “It’s a certainty that SPARK will enable new SFARI is also supporting research into ‘somatic’ mutations GENES risk-gene discovery in autism,” says Wendy Chung, the — ones that occur during development rather than at or initiative’s principal investigator. before fertilization, and so appear in only a fraction of an individual’s cells. Several papers in the past year indicate Collecting genetic data from 50,000 families should not that somatic mutations contribute significantly to only reveal new rare autism mutations, but also jump-start autism risk. the search for the common gene variants that affect autism risk — variants that, by definition, appear in at least 1 Between 5 and 10 percent of children with ‘simplex’ autism percent of the general population. These common variants — autism that affects no one else in their family — may are often harmless but can be damaging when joined with have one of these somatic mutations, says Brian O’Roak of just the right combination of other common variants. “We know that collectively they impart a significant Between 5 and 10 percent of risk for autism, but for the most part we don’t know children with ‘simplex’ autism which ones they are yet,” Reichardt says. may have somatic mutations. Searching for common variants requires tens of thousands Oregon Health and Science University in Portland, whose of samples. “If we get to 50,000, that dramatically changes team, with support from SFARI, has been combing the what can be done,” Reichardt says. “To date we have not SSC for somatic mutations. So far, his research group has been able to find out much about common variants, but our uncovered somatic mutations both in known autism risk knowledge is going to accelerate over the next year or so.” genes and in novel genes that had not previously been connected to autism, but that are involved in biological Diving deeper pathways linked to the disorder. Although sequencing tens of thousands of SPARK families is expected to be a key driver of future discoveries, there’s Although somatic mutations may sound more benign

A visualization of the human genetic information contained in the SFARI Gene database. The outer shell displays a curated subset still a lot to learn from the roughly 2,600 families in than mutations that appear in every single cell, it’s possible of genes spread across the 23 chromosome pairs. The color of each gene indicates the confidence of the gene’s predicted link to the Simons Simplex Collection (SSC), a data repository the reverse is true. “Some of the somatic mutations we’re autism, with red indicating the strongest links. The shell’s interior connects genes associated with the same protein interaction. launched by the Simons Foundation in 2007. Many of the finding might be so bad that if they were in every cell, An interactive version of the visualization is available at gene.sfari.org/database/ring-browser. early sequencing studies of the SSC focused on the exome, that would not be compatible with life,” O’Roak says. where it is most straightforward to find direct links between Sequencing SPARK families should clarify the role of Over the past five years, sequencing studies have made their role in autism unmistakable. mutations and autism, but the exome represents only about these mutations, he says. of individuals with autism and their families Yet researchers estimate that 300 to 1,000 1.5 percent of the human genome’s 3 billion nucleotides. have led to the discovery of about 100 high- genes may confer risk for autism. To shake out With support from SFARI, the New York Genome Center All told, the mutations identified so far by all methods confidence autism risk genes — a remarkable these additional genes, the Simons Foundation has completed whole-genome sequencing of the families explain about 20 to 25 percent of simplex autism, Packer step forward in understanding the genetic Autism Research Initiative (SFARI) is pursuing in the SSC. Researchers hope this sequencing data will says. “It’s fair to say that this number will only grow, and basis of the condition. These studies have a wide range of approaches. illuminate how the non-protein-coding regions of the probably fairly rapidly.” successfully gleaned many of the mutations genome — such as regulatory regions, which control most prominently involved in autism: the ones where and how much a given gene is expressed — that, though rare, appear frequently enough to affect autism risk. SFARI

26 SIMONS FOUNDATION A light switch Changing the balance of excitation and inhibition in neurons instantly reduces social deficits in a mouse model of autism, researchers have found. The finding, by SFARI Investigator Karl Deisseroth and others at Stanford University, implies that the wiring in the brains SFARI RESEARCH of these mice is still capable of relatively normal social functioning, and suggests that treatments that change Antibiotic for autism? the excitation-inhibition balance might alleviate social Maternal exposure to infection during pregnancy raises the risk that the difficulties in some individuals with autism. baby will have an autism-related condition, according to epidemiological ROUNDUP studies. And mice exposed to a maternal infection in utero have a The researchers, who reported their findings August 2, higher risk of autism-like behaviors. But pre-treating pregnant mice 2017, in Science Translational Medicine, engineered mice with an antibiotic protects their pups from developing these behaviors that lack CNTNAP2, a gene linked to autism, and also after a maternal viral infection, even though antibiotics don’t kill express special proteins that turn neurons on or off in viruses, researchers reported in the September 28, 2017, issue of response to a flash of light. Nature. The study was led by SFARI Investigators Jun Huh of Harvard University, Dan Littman of New York University and Gloria Choi of the Preferred parents Mice lacking CNTNAP2 normally show no interest in Massachusetts Institute of Technology. People and animals have two copies of most genes — one interacting with unfamiliar mice. But when the researchers copy from each parent — and conventional wisdom says that exposed these mice to a flash of blue light that caused The study suggests that infection-induced autism-like features stem Over the past 15 years, the Simons Foundation Autism Research Initiative the two copies are expressed equally in the body. Yet studies inhibitory neurons in the prefrontal cortex to fire, the mice from inflammatory immune responses produced by maternal gut (SFARI) has supported more than 400 researchers studying autism and of cultured cells have indicated that this may not always be became interested in strange mice placed in their cage; the bacteria. Previous studies by several of the investigators who worked on related disorders. In 2017 alone, SFARI provided funding to more than the case. Now a study published March 8, 2017, in Neuron same thing happened when the team used flashes of light the Nature paper showed that in pregnant mice with a viral infection, 200 investigators from around the world, who have studied everything has shown for the first time that this conventional wisdom to turn off excitatory neurons. The flashes of light also make inflammatory molecules made by ‘T helper 17’ immune cells contribute from sex differences in a mouse model of autism to the role of maternal is not true in the brains of mice, monkeys and humans. the mice less hyperactive, the team found. These results to the development of cortical abnormalities, social deficits and repetitive gut bacteria in the condition. The following are some highlights of the boost the long-held theory that over-excitement in the brain behaviors in the mice’s pups. research of SFARI Investigators in 2017. SFARI Investigator Christopher Gregg, of the University is responsible for many autism symptoms. of Utah in Salt Lake City, and his collaborators found The research may help explain why some Making sense of missense that mouse livers and muscle tissues and a brain region Researchers have uncovered 200 candidate risk genes for autism and called the dorsal raphe nucleus all have some genes that Sex differences in learning mutations linked to autism affect certain other neurodevelopmental conditions by examining ‘missense’ mutations preferentially express a particular parent’s copy. When Male mice with an autism-related mutation people more strongly than they do others. — mutations that change only one nucleotide in a gene, but in a way that the researchers mutated one parental copy of one of these struggle with learning tasks that females with alters the resulting protein. Even though missense mutations are thought genes, they found patches of expression of the mutated gene the mutation can perform as well as controls, a new study In the new study, the researchers gave pregnant mice vancomycin, an to account for more cases of autism than more damaging mutations throughout the mouse brain. The team also found genes has found. The work, led by SFARI Investigator Ted Abel of antibiotic that kills the gut bacteria that spur the growth of T helper called ‘likely gene-disruptive’ mutations, they have been studied far with parental preferences in macaque monkeys and in the University of Iowa, illuminates some of the mechanisms 17 cells. Even though the mice were exposed to a virus after antibiotic less, partly because they are so common in the general population that postmortem human brains. that appear to protect girls from autism. Autism is almost treatment, their pups developed normally, the researchers found. disentangling their role in autism is tricky. five times more common in boys than in girls. Several of the genes identified in macaques are associated Certain bacteria in human intestines also promote the growth of T helper Now, a team led by SFARI Investigator Evan Eichler of the University with autism, and others are linked to bipolar disorder, The mice in the new study, which was published October 17 cells. Mice colonized with these bacteria and then infected with a of Washington has analyzed missense mutations in 8,477 people with intellectual disability or other neurodevelopmental 17, 2017, in Molecular Psychiatry, lacked a stretch of DNA virus gave birth to pups with autism-like features, the researchers found, neurodevelopmental conditions, along with 2,178 controls. The team conditions. Among the autism genes, one of two parental called 16p11.2. About 30 percent of people with a deletion unless the mothers were pre-treated with an antibody that blocks an found 200 genes that have significantly more missense mutations copies of DEAF1 is also preferentially expressed in four in 16p11.2 have autism, and mice with the deletion have inflammatory molecule the T helper 17 cells produce. The researchers in individuals with a neurodevelopmental condition than in controls; human brain regions, the team found. The research may cognitive deficits and are hyperactive. hypothesize that women whose gut microbial community is skewed 79 percent of these genes have never before been associated with a help explain why some mutations linked to autism affect toward the bacteria that generate T helper 17 cells may be more likely neurodevelopmental condition. One of the most frequently appearing certain people more strongly than they do others: The Male mice with the deletion take about three times as long than other women to have children with autism if they get an infection missense mutations was found in a gene called GRIA1, which encodes mutation’s effect may depend on which parent it came from. as females with the deletion and controls to learn to poke during pregnancy. a receptor for a neurotransmitter called glutamate. their noses into a hole for a reward. Males with the deletion also show less perseverance than other mice when the The study’s techniques, described in the August 2017 issue of Nature task gets harder. Neuroscience, offer a way not just to identify candidate risk genes, but also to pinpoint just which parts of a gene confer the most risk if they Abel and his colleagues examined the striatum — one of are mutated. This information may prove valuable down the road the brain’s reward centers — in the mice and found that when it comes to designing targeted therapies for autism and other compared with controls and female mice with a 16p11.2 neurodevelopmental conditions. deletion, male mice with the deletion have increased activity in a signaling pathway that dampens neuronal activity, and also increased expression of a dopamine receptor that inhibits learning. Abel is planning further work to drill down into the molecular mechanisms underlying this vulnerability in males and resilience in females. SFARI 28 SIMONS FOUNDATION SPARKING AUTISM Autism researchers can now apply to SPARK’s research match program, which connects them with appropriate SPARK families for RESEARCH their study. In its first year, the program helped researchers recruit thousands of families for more than 10 studies, ranging from assessments of environmental exposure during pregnancy to clinical investigations Researchers who want to study autism often face a big of brain function. obstacle before they even get started: how to track down enough individuals and families who meet the criteria for “It’s so exciting to see how engaged SPARK their studies, and then gain their participation. To overcome participants are and how they are accelerating this obstacle, in April 2016 SFARI launched SPARK the pace of research,” says Pamela Feliciano, (Simons Foundation Powering Autism Research for SPARK’s scientific director. Knowledge), an initiative that aims to create a repository of Around 10 percent of SPARK participants will receive a result about the genetic cause of genetic and behavioral data from 50,000 individuals with In addition to providing a platform for their autism, if they elect to receive it. “I had a million questions until we participated in the SPARK study and got the genetic finding,” says Lynn Vigo, left, a mother of an adult autism and their families. The effort is well on its way. researchers to connect with potential study with autism. “The finding explained so much of what we struggled to understand.” participants, the initiative is also in the process Two years into the initiative, SPARK of genetically analyzing the saliva samples, In addition to returning genetic results to has made substantial progress which will vastly expand the pool of genetic families, SPARK is starting to provide families toward its target: more than 39,000 data from families with autism. “SPARK with individualized feedback that shows individuals with autism have families’ data will help power a new level of how their child’s social questionnaire score already enrolled. “There has been discovery,” Feliciano says. compares with that of other children an outpouring of support from the in SPARK, and it also hosts a series of autism community,” says Wendy Chung, the To date, SPARK’s genetic analyses have webinars that have attracted more than 4,500 initiative’s principal investigator and focused on the ‘exome’ — the protein-coding attendees. “An important goal of SPARK is SFARI’s director of clinical research. regions of the genome — because those are to empower participants with knowledge the regions in which it is easiest to establish a and access to experts,” Feliciano says. “As a SPARK participants complete several clear causal connection between mutations and parent of a child with autism myself, I know online questionnaires and mail in autism. SPARK has completed whole-exome firsthand how helpful a little bit of data or a saliva samples from themselves and sequencing for 488 families and has made the new insight can be. We aspire to arm all family members. They then have the data freely available to scientists. An additional participants with information that will be option of taking part in future research 4,011 families are in the pipeline for whole- useful to them.” studies, although such participation exome sequencing.

A recruitment video delineates the SPARK participation process and underscores is never mandatory. the mutually beneficial partnership of SPARK participants and autism researchers The initiative maintains a list of high- in understanding autism. www.sparkforautism.org To boost enrollment from fathers, who are underrepresented confidence autism risk genes, and as families among SPARK participants, the initiative ran a three-month- are sequenced, they have the option of being long outreach campaign called “Men of Action,” starting notified through a physician or genetic on Father’s Day 2017. “This was our first attempt to really counselor if they have a mutation in one of connect with dads, and it got a very positive response,” these high-risk genes. So far, about 5 percent says Amy Daniels, SPARK’s project manager. “We saw a of families have been eligible to receive results, notable increase in the number of fathers who enrolled a proportion that is likely to grow as new and sent in saliva samples.” high-confidence genes are identified. SPARK has already expanded its high-confidence gene list by about 15 percent in the past year, Feliciano says. SFARI

30 SIMONS FOUNDATION One of the most immediately apparent traits of many individuals with autism is their diminished eye contact during social interactions. Warren Jones and Ami Klin, researchers at the Marcus Autism Center at Emory University in Atlanta, have spent more than 15 years teasing out the ways in which the gazing preferences of people with autism differ from those of the general population. The duo explores how people on the spectrum look at and learn about the social world — and how differences in these gazing behaviors manifest early in life. And in 2017, in a collaboration with John Constantino of Washington University in St. Louis, they uncovered compelling evidence that gazing differences are tightly tied to genetics.

Many of Jones and Klin’s deepest insights into autism conserved and foundational mechanism of socialization. have emerged from the unusual and ambitious design of And in 2013, the team reported that gazing differences TRACKING their experiments: Instead of studying children who have appear even earlier in development: Baby boys who are already been diagnosed with autism, as most researchers later diagnosed with autism start losing interest in looking at eyes sometime between 2 and 6 months — the earliest behavioral marker of autism found to date. This marker is “This is really how you predictive of a diagnosis of autism, and severity of features, TWINS at 36 months, they found. should be doing science.” Most recently, Jones, Constantino, Klin and their colleagues reported July 20, 2017, in Nature that the way infants view do, they focus instead on those children’s baby siblings to social scenes is highly influenced by genetics. Identical try to spot the behavioral differences underlying autism at twin toddlers, they found, make strikingly similar decisions the earliest possible age. ‘Baby sibs’ of children with autism about not just what to focus on but even how to shift their are 20 times as likely as infants in the general population to gaze in search of social information from moment to eventually be diagnosed with autism. moment. “Identical twins effectively synchronized their looking at social content,” Jones says. “We found remarkably Such experiments are difficult and expensive, as not only do strong evidence that genes directly shape the way a child they typically take years, but they also must include far more sees the world.” children than will ultimately receive an autism diagnosis. Even though baby sibs have an elevated risk of autism, only The types of gazing that the team found to be most about one in five of them eventually receives a diagnosis. powerfully influenced by genetics — attention to eyes and mouths — are the same ones that are strongly “This is really how you should be doing science,” says John decreased in children with autism. “It’s gratifying to Spiro, deputy scientific director of the Simons Foundation see Warren and Ami’s early work extended in this really Autism Research Initiative. “But it’s very hard to carry out.” interesting way and starting to make connections to the genetics of autism,” Spiro says. The Simons Foundation was a key early funder of Jones and Klin’s work, providing support both for specific studies (including 2009 and 2013 papers in Nature) and for the research infrastructure that made the team’s long-term studies possible. “That platform has enabled some of our most important insights into the early development of autism,” Jones says.

For instance, it allowed them to figure out in 2009 that, unlike typical toddlers, 2-year-olds with autism have no preference for ‘biological motion,’ an evolutionarily highly

New research on twins indicates that the way children look at the social world — what they look at, when and for how long — is strongly influenced by genetics. Image courtesy of John Constantino of the Washington University School of Medicine in St. Louis SFARI

32 SIMONS FOUNDATION “A place like this can be a game-changer because it can really open up what learning is about,” says Honey. “Every learner can find a foothold here that makes them feel successful. We can create experiences here that are designed to ensure that every visitor feels a sense of possibility, a sense of ‘I can do this.’”

SCIENCE SANDBOX: A third Science Sandbox partner, the Exploratorium, Science Sandbox also partners with non-museum organizations that take focuses on helping people learn on their own through similarly innovative approaches to communicating science widely. interactive exhibits. Founded by physicist and educator THE CHANGING FACE Frank Oppenheimer in 1969, this highly regarded public The Science Festival Alliance’s Just Add Science initiative also meets science museum and learning laboratory in San Francisco, people on their own turf, infusing science into venues such as county California, encourages visitors to be curious and ask fairs, sporting events and Renaissance festivals. Pioneer Works, a cultural OF SCIENCE MUSEUMS questions. In the museum’s Tinkering Studio, for example, center in a restored warehouse in Brooklyn’s Red Hook neighborhood, visitors use tools such as pliers, scissors and sewing needles offers programming such as the Scientific Controversies series, which to create projects — such as wearable working circuits — features speakers such as biologist Richard Dawkins and Nobel laureate that they can follow up with at home. Rainer Weiss in conversation about provocative topics in science. And the Some people aren’t comfortable in big, formal science San Francisco Bay Area’s Science Action Club, designed by the California museums. Others face economic barriers to entry or don’t The museum also creates science experiences for the people Academy of Sciences, is a nationwide out-of-school program for middle have such an institution in their community. Luckily, getting outside its walls: its Studio for Public Spaces places exhibits schoolers that encourages curiosity about the natural world while involved in science is becoming more and more apt to read the copy and share information. After exploring the throughout San Francisco. When people encounter an providing students with hands-on activities that promote STEM learning. happen outside a formal setting. Evolving explorations of museum side by side, people who just met sometimes walk exhibit or a novel, surprising environment while going about what it means to be a modern science museum means that away in conversation, Schochet says. One visitor left the their daily lives, this can create a welcome interaction and All these programs further Science Sandbox’s goal of helping people many more people are encountering — and participating in museum announcing: “I’m a mollusk person now!” learning experience. Some displays foster examination of realize that they do not have to be a scientist to think like one. “We — scientific programming, wherever they may be. the physical senses, whereas others bring people together believe society as a whole benefits when people engage in things like “It’s important for everyone to have access to this kind of to explore social interactions. Engaging with people outside evidence-based reasoning, informed decision-making and critical Science Sandbox, a Simons Foundation initiative whose knowledge, and for it to be presented in ways that someone of the museum often creates a more meaningful experience thinking,” says Greg Boustead, program director for Science Sandbox. mission is to unlock scientific thinking by engaging people with no background in science can jump in and begin for them, says Robert Semper, the museum’s associate “It’s important to encourage people to be active participants in science, with the process of science, is now helping some of its to see the big picture,” says Schochet. “The dream is to executive director. not just tell them that science is important. When they make their own grantee partners upend the stereotype of the traditional, install 6-foot-tall museums all over the country. Even in the connections — through their own context and their own background — staid science museum. Department of Motor Vehicles!” “For people who live in this world, in this society, it’s we think it’s a powerful way to engage.” critically important for them to both have the skills and the MICRO, a nonprofit founded in 2016 by computational Another Science Sandbox partner, the New York Hall of experience to be independent learners,” Semper says. “This ecologist Amanda Schochet and media producer Charles Science (NYSCI) — located in Queens — uses science to means providing learning experiences where people live and The Exploratorium installed this public interactive exhibit in San Francisco, Philipp, works with designers, artists and scientists to pack bridge cultural and economic divides. The museum’s NYSCI work day to day, as well as providing them in more formal called ‘Whispering Dishes.’ A pair of curved concrete dishes (just one is shown) focuses sound so that two people around 15 meters apart can hear knowledge into cabinets of curiosity about the size of a Neighbors program, an outreach effort in Queens’ Corona settings like museums.” one another whispering, even over the din of a busy street. vending machine. MICRO increases exposure to science neighborhood, launched Science Ambassadors in February Image courtesy of the Exploratorium learning by partnering with venues such as the Brooklyn 2017. The museum opens up after school, free of charge, to Public Library, Ronald McDonald House and Rockefeller students in a network of 20 collaborating schools within Center to install these tiny museums. The first museum in walking distance. Children get help with homework, explore their ‘fleet,’ the Smallest Mollusk Museum, can go almost exhibits in the museum and take advantage of the Design anywhere, from hospital waiting rooms to community Lab and Maker Space. centers and malls, reaching people where they already are. Many schools in the area — where most residents are The mollusk museum features a 3-D-printed octopus brain immigrants — don’t provide much access to STEM and a glowing holographic aquarium, where visitors can programming, says Margaret Honey, president and chief peek inside to see virtual images of mollusks swimming, executive officer of NYSCI. And if you are an undocumented squid hunting or octopuses squiggling around. Continue immigrant, a large museum might be all the more around the display to learn how the creatures live, how intimidating. So NYSCI Neighbors hopes to build trust similar they are to aliens in movies and how they’ve within Corona by offering resources and shared experiences adapted to survive. that help both parents and children engage in science.

Not only does the Smallest Mollusk Museum teach people “I think it’s a good opportunity for kids and parents to about the mollusk world, its compact size physically brings come in and then to get exposed to the rest of the stuff visitors closer together as they learn. While children study that’s in the museum,” says Luz Salazar, whose daughter, the more visual aspects of the exhibit, their parents often Camila Melendez, gets homework help through Science Ambassadors. When Camila finishes her studies, she takes part in design-make-play classes that nurture creative problem-solving. OUTREACH AND EDUCATION

34 SIMONS FOUNDATION have during the school year — and something that would bridge the and communities. Overall, the organization works to inspire, divide between science and math teachers.” support and share new ideas that shape policy, practice and the retention of excellent teachers in mathematics Summer Think sessions included a ‘deep dive’ into how to incorporate MATH FOR and science across New York City’s 1,700 public schools climate change justice into lesson plans by combining statistics and and beyond. climate science. Led by MƒA Master Teacher Peter Mulroy, the session explored analyzing demographic and climate datasets to uncover links AMERICA: MƒA awards renewable four-year fellowships, which between poverty, race and future climate impacts. “It was nice to go include a stipend and frequent opportunities for teachers beyond just scatter plots related to climate justice,” Ginsberg says. to connect with one another. The organization also offers Teachers who attended the session are still swapping materials and extensive professional learning and growth opportunities, ideas online, she says. SUMMER such as courses, workshops and seminars, throughout the school year. Engaging students with chemistry sometimes means swapping beakers for spatulas. THINK The original spark that led to Summer Think was anything Math for America (MƒA) Master Teachers but straightforward. It came from MƒA Master Teacher Hayeon Rachel Jun and Laryssa Kramarchuk Brian Palacios, a high school math teacher at the Bronx blend science and the culinary arts in their Center for Science and Mathematics. Palacios received a classrooms. Lessons have students hand- grant from MƒA to attend a summer meeting of teachers churning ice cream, growing rock candy and That session, and many others like it during the meeting, who collaborate on Twitter. In his written reflection on that quantifying the hotness of peppers, all while inspired Summer Think participants to try new approaches meeting, he proposed that MƒA host a summer conference. studying the underlying scientific properties themselves. Jun and Kramarchuk “were so honest about MƒA staff jumped on the idea and recommended that of these materials. their experience and how it had gone, including the things Palacios spearhead the planning, with their support. “It was that didn’t go well,” says fellow MƒA Master Teacher just a crazy idea,” he says. “I didn’t have the slightest clue Jun and Kramarchuk shared these experiences Courtney Ginsberg, who teaches high school mathematics how to plan a conference. I didn’t even know how to start.” with fellow math and science teachers last July at Humanities Preparatory Academy. “The session made at a three-day MƒA conference called “Summer me feel more comfortable to do something seemingly crazy After sending out feelers for other MƒA teachers interested Think.” Their session, “Kitchen Chemistry,” like that in my classroom, where it could be a mess of kids in helping plan the as-yet-unnamed conference, things was one of many opportunities at the MƒA throwing ice cream around.” remained uncertain for a time, Allison says. “We wondered Summer Think for teachers to collaborate and if a large enough group would be interested in a summer share their ideas on how to engage students The 2017 Summer Think was MƒA’s first summer event. Twenty teachers showed up for the first planning with mathematics and the sciences. conference. “You can feel the difference in a room full meeting, which was too many to plan a conference, so they of teachers when it’s summer,” says Courtney Allison, got some ideas and made a smaller planning committee.” This inaugural summer series at MƒA’s deputy executive director for MƒA. “Everyone was a lot more New York City headquarters continued relaxed and casual. Teachers repeatedly tell us how much Leah Hirsch, a program officer at MƒA, shepherded the the organization’s mission of supporting they love when we do things outside of the normal school teachers through the uncertainties of the process. “We Master Teacher Jeffrey Horenstein discusses his group’s paper craft design during a ‘deep dive’ session facilitated by fellow Master Teacher Vielca Anglin. outstanding teachers, which it does by fostering year because they have time to think. You’re not in the provided the structural support, but the teachers came up The Summer Think session, titled “Curriculum Reboot: Design Challenge as collaboration, offering professional growth thick of things.” with the conference ideas and all of the programming.” Assessment,” explored how integrating math and science into the design process opportunities and providing financial support can inspire empathy, critical thinking and real-world problem-solving in students. to teachers. Importantly, all the sessions at This type of support is what MƒA does — and it means Summer Think are led by and for teachers. “We provided the structural support, but the a lot — says Ginsberg, who also served on the planning committee. “MƒA had trust in us as teachers to put together Other sessions went beyond curricula. “Facilitation as Leadership” Jun and Kramarchuk demonstrated one teachers came up with the conference ideas and fund this conference,” she says. “It feels good to be part discussed how teachers can embrace their role as leaders to ensure equity of their chemistry lessons about phase and all of the programming.” of a community where they trust us and believe that we have and access for all students. Attendees assembled a toolkit of ways to transitions and crystallization. Attendees the best intentions at heart, which is not something we all manage group dynamics and give students a voice. The session helped shook plastic baggies of salt, ice, sugar and John Ewing, president of MƒA, says the organization is get in our school communities.” build strong connections among teachers, Lennon says. cream, periodically pausing to measure the no stranger to doing things differently. “MƒA was begun temperature of their solidifying ice cream with a revolutionary idea — that the best way to improve With MƒA’s help, the planning team gathered proposals “We had to let our guard down,” she says. “At MƒA, I feel like I can do that. mixes. Whereas pure water freezes at 32 teaching is to focus first on the most accomplished teachers. for sessions. “We got a bunch of proposals, and so we At the Summer Think, we shared one of the things we’d like to improve degrees Fahrenheit, the addition of salt and Excellence is best built on excellence.” made our own rubric and graded them,” says MƒA Master about our teaching or that we didn’t think was going well. It’s hard as a sugar lowers the mixture’s freezing point by Teacher Diana Lennon, an environmental science teacher teacher to put yourself out there and say, ‘I don’t think I do this well.’” a few degrees, as the salt and sugar molecules MƒA’s goals are to keep the most accomplished math and at Columbia Secondary School, who also served on the get in the way of water molecules joining science teachers in the classroom, foster teacher-to-teacher planning committee. “We were looking for proposals that Facing uncertainties and trying new things as a community were to crystallize. professional growth, provide teachers with opportunities were outside of the things we could do during the time we common themes throughout the conference, from the session topics to for leadership, and influence and build a professional the meeting’s very planning. And they’ll be central themes for the 2018 community of accomplished teachers to collaborate and Summer Think, too. “There were a lot of people trying things for the learn from one another. MƒA now supports more than first time at this conference,” says Ginsberg. “It’s important for teachers 1,000 math and science teachers in New York City who are to step into that role, because that’s what we’re always asking our making a lasting impact on their students, schools students to do.” OUTREACH AND EDUCATION

36 SIMONS FOUNDATION The Simons Foundation’s research and grant-making divisions are spearheaded by accomplished scientists. SCIENTIFIC These individuals bring expertise, experience and creative thinking to overcoming the uncertainty involved in LEADERSHIP advancing research in mathematics and the basic sciences.

Ian Fisk

Ian Fisk is co-director of the Flatiron Institute’s Scientific Computing Core, which manages the institute’s computational infrastructure and Antoine Georges Gerald D. Fischbach develops computational software for Nick Carriero the scientific community. “We try to Antoine Georges is director of “The human brain is the most make sure that scientific progress Gregory Gabadadze the Center for Computational “Ours is a world in which we try complex and mysterious object in is limited by how quickly the people Quantum Physics (CCQ) at the to beat out all uncertainty,” Nick the known universe,” says Gerald D. can think and understand, and Gregory Gabadadze joined the Flatiron Institute and a pioneer Carriero says. Carriero co-directs Fischbach, distinguished scientist not by how quickly computers can Simons Foundation in 2017 as of dynamical mean-field theory, a Marian Carlson the Flatiron Institute’s Scientific and fellow at the Simons Foundation. process and access data,” he says. associate director for physics in the method to determine the electronic Computing Core, which develops, “Once we understand even a hint Mathematics and Physical Sciences structures of materials containing Marian Carlson is director of the deploys and maintains the institute’s about how nerve cells in the brain Fisk majored in mathematics division. He is chair of the physics electrons that display collective, Simons Foundation’s Life Sciences computational infrastructure and give rise to thoughts, emotions and and physics in college before department at New York University rather than individual, behavior. division. The division supports, develops software for the greater behavior, it will break open a whole earning his M.S. and Ph.D. from and served as director of the alongside more traditional awards scientific community. new world.” the University of California, San university’s Center for Cosmology At the CCQ, Georges continues programs, collaborations devoted Diego. He later served in several and Particle Physics. his innovative research into the to the study of marine microbial Coders, he says, expect the same Fischbach first wondered how the roles overseeing the preparation of behaviors of the unfathomable ecosystems and the origins of life. chunk of code to function the mind works while taking philosophy the computational infrastructure In high school, he was inspired numbers of electrons that govern These research areas contain “huge same way even on different courses in college. Seeing few for the Compact Muon Solenoid by an article that explained that a the properties of molecules and unknowns and great complexity,” she hardware. Carriero ensures that the graduate programs in neuroscience, experiment within CERN’s Large vacuum isn’t truly empty but instead materials. The quantum mechanical says. The rock record from the time institute’s computers and software he decided to attend medical school. Hadron Collider program. He is populated by fleeting virtual waves behavior of all those electrons is life arose on Earth has been erased perform reliably, drawing from He went on to Harvard University, joined the Flatiron Institute in 2014, and particles popping in and out of “simple to formulate but extremely almost entirely, she says, and marine his experience managing high- where he studied the formation then called the Simons Center for existence due to quantum uncertainty. hard to solve,” Georges says. microbial communities across the performance computing platforms and maintenance of synapses, Data Analysis, because building up That germ of wonderment led him to Together with co-director world contain incredible diversity. as co-director of Yale University’s which play an essential role in the computational infrastructure earn a Ph.D. from Rutgers University, Andrew Millis and distinguished W.M. Keck Biotechnology Resource communication in the brain. His for a multidisciplinary research after which he began researching the research scientist Angel Rubio, he Carlson’s interest in the life sciences Laboratory. “If you run the same expertise led to professorships at institute “sounded like a great universe at the smallest and largest leads the center’s efforts to develop comes from a lifelong fascination thing designed to be deterministic Washington University and Harvard challenge,” he says. scales by studying particle physics, ideas, algorithms and code to with the outdoors and the biosphere. twice and you get different answers, Medical School. He subsequently gravity and cosmology. help tame the complexities of the She majored in biochemical that’s a bad thing,” he says. served as director of the National The core’s work, he says, involves quantum world. sciences, examining living cells at Institute of Neurological Disorders adapting to the evolving challenges His work, he says, is motivated the molecular level, and earned her Sometimes a bit of uncertainty can and Stroke and then as executive of supporting scientific research. by the big outstanding questions Georges received his Ph.D. from Ph.D. from Stanford University. help drive discovery, though. “Non- vice president for health sciences at “How people need to work to make in physics. “We don’t understand the École Normale Supérieure in She is professor emerita of genetics determinism and randomness Columbia University. progress in their fields is always about 95 percent of what our France and is professor of physics and development at Columbia are cousins of uncertainty,” he changing,” Fisk says. “How to universe is made of,” he says. “We at the Collège de France, where he University and served as senior says, “and they both play important In 2006, Fischbach joined the plan for capacity and capability is don’t understand the patterns of holds the chair in condensed matter associate dean and vice dean for roles in computing.” Computers foundation as founding director constantly changing.” masses and interactions of the most physics. He is a member of the research at the medical school. Her can generate random numbers of the Simons Foundation Autism elementary particles that make the French Academy of Sciences and laboratory used genetic analysis that can help assess an algorithm’s Research Initiative (SFARI). things around us.” Those puzzles, shared the 2006 Europhysics Prize to identify and study conserved efficiency or broaden a simulation’s As a founder of the Simons he says, “keep me up at night.” in condensed matter physics for his mechanisms of signal transduction scope. “There’s no point running Collaboration on the Global Brain, contributions to the development and transcriptional regulation. a simulation multiple times if it’s Fischbach says, “The Global Brain of dynamical mean-field theory. deterministic. But by introducing studies internal mental states; randomness, your simulation will SFARI studies how the brain explore something slightly different develops. I am optimistic that these each run,” Carriero says. projects will enhance one another.” SCIENTIFIC LEADERSHIP

38 SIMONS FOUNDATION Andrew Millis Yuri Tschinkel

Leslie Greengard Andrew Millis is co-director of the At age 15, Yuri Tschinkel picked Center for Computational Quantum up a book about number theory, “Computational biology is an Physics at the Flatiron Institute. which sparked his enduring interest interesting intersection of Alex Lash He formerly served as associate Louis Reichardt in the theory of numbers. That problems that come from biology director for physics at the Simons interest led him to earn a Ph.D. with methods that come from Alex Lash is the Simons Foundation’s Foundation. Millis is also professor Louis Reichardt is director of from the Massachusetts Institute of mathematics,” Leslie Greengard says. chief informatics officer. His of physics at Columbia University the Simons Foundation Autism Technology and pursue a career as team helps manage, analyze and and director of the Simons Research Initiative (SFARI), David Spergel a research mathematician, working Greengard has, fittingly, both an disseminate the foundation’s Collaboration on the Many Electron overseeing the group’s efforts to at the interface of algebraic and M.D. and a Ph.D. in computer extensive scientific datasets and Problem. In July 2017, he received study the causes of autism and David Spergel directs the Center for arithmetic geometry. Now, as science from Yale University. develops software tools to advance the Hamburg Prize for Theoretical enhance the quality of life of Computational Astrophysics at the director of the Simons Foundation’s He is director of the Center for the foundation’s mission. The work Physics for his contributions to the individuals with the condition and Flatiron Institute. The center, he Mathematics and Physical Sciences Computational Biology (CCB) at combines his training in anatomic field of condensed matter physics. their families. says, aims to “address some of the (MPS) division, he is dedicated to the Flatiron Institute and Silver pathology and a lifelong love of basic questions that we have about supporting mathematics, physics Professor of Mathematics and computers. He was previously Millis’ research focuses on Reichardt initially intended to the universe: How did it begin? and theoretical computer science on Computer Science at New York a laboratory member in the understanding how interactions study history in college but was How will it end? How did the Earth, a much broader scale. University. Together with Vladimir computational biology program between particles influence the inspired by the Sputnik satellite, stars and our galaxy form? What Rokhlin, he developed the fast and then scientific director of properties of molecules and molecular biologist James Watson, makes up the universe?” Conducting mathematical research multipole method, a mathematical the bioinformatics core facility at materials. “The intellectual challenge and novelist and physical chemist and overseeing MPS each provide technique with a wide range of Memorial Sloan Kettering Cancer is to understand the quantum C.P. Snow to change his major Spergel has spent his career unique challenges, he says. “In applications that speeds up the Center in New York. mechanics of interacting many- to biology. After attending the tackling some of those cosmological number theory, there is no calculation of long-range forces in particle systems,” he says. “You put University of Cambridge as a conundrums. Since receiving his uncertainty,” Tschinkel says. “In the n-body problem. The Institute of Designing a software tool requires large numbers of particles together, Fulbright scholar and completing Ph.D. from Harvard University, he decisions on grant applications, Electrical and Electronics Engineers a firm understanding of how people and they do crazy and interesting his Ph.D. at Stanford University, he has focused on probing the nature there is a lot of uncertainty: Panels named the method one of the top 10 will use the system, Lash says, but things. We want to understand how changed his research interest from of dark matter and uncovering new sometimes produce divergent algorithms of the 20th century. developing that understanding can and why this happens.” genetics and molecular biology physics. He played an influential opinions; long-term projects may or be tricky. “There’s always a bit of to neuroscience as a postdoctoral role in the Wilkinson Microwave may not lead to exciting science.” As director of the CCB, Greengard uncertainty about what users want, Ultimately, Millis hopes, the fellow at Harvard University. Anisotropy Probe project, which oversees groups working on even if you have them right in front center’s research will help materials mapped temperature fluctuations in Tschinkel is excited by the utility that modeling frameworks, algorithms of you. They can’t perfectly foresee scientists not only understand the At the University of California, the heat left over from the Big Bang. computational tools are bringing and software to help scientists probe how they might use the system, and properties of known compounds, San Francisco, his lab identified to areas of pure mathematics. Such large experimental datasets for new neither can we,” he says. but also design new materials and the protein synaptotagmin 1, later In December 2017, Spergel methods allow mathematicians to insights into the biological world. molecules with desired properties. shown by others to be critical and his colleagues received the tackle problems that are larger and Lash’s team addresses this challenge “Thanks to a remarkable combination for neurotransmitter release. He 2018 Breakthrough Prize in more complex than ever before. by putting themselves in users’ of new experiments, new theoretical completed many studies on Fundamental Physics for their But even with such technological shoes during pre-launch designing and computational methods, and neurotrophins and their receptors, work on that project. In addition advances, “sometimes one needs and testing, and then by tracking the unique environment provided which are essential for neuronal to his directorship, he is Charles A. just a brilliant idea” to make a feedback after a launch and fine- by the Flatiron Institute, we have survival and function. By combining Young Professor of Astronomy at breakthrough, he says. tuning the software accordingly. the chance to really make progress genetic, cellular and biochemical Princeton University. “That’s how you make your users on fundamental understanding and analyses, his group characterized the happy,” he says. control of the quantum properties of impact of cell adhesion molecules on materials,” he says. autism-related behaviors. In July 2013, he assumed directorship of SFARI. SCIENTIFIC LEADERSHIP

40 SIMONS FOUNDATION FINANCIALS BALANCE SHEET

Assets 12/31/17* 12/31/16

Cash and Cash Equivalents 247,842,261 160,564,011 Investment Portfolio 2,800,626,072 2,663,512,864 Autism Flatiron Institute Property and Equipment, Net 246,719,463 206,583,733 15.39% 0.25% Other 2,028,544 2,360,299

Total 3,297,216,340 3,033,020,907

Outreach and Mathematics & Physical Liabilities 12/31/17* 12/31/16 Education Sciences 15.14% 31.77% Accounts Payable 17,157,197 13,505,532 Grants Payable 545,446,555 524,286,998 Capital Lease Obligation 151,135,182 147,834,375 Deferred Rent 2,852,656 3,621,571 Deferred Excise Tax Liability 20,638,967 20,638,967

Total 737,230,557 709,887,443

2017 Grant Payments Net Assets 2,559,985,783 2,323,133,464 by Category INCOME STATEMENT

Life Sciences For 12 Months Ended For 12 Months Ended 37.45% Revenue 12/31/17* 12/31/16 Proportions of Expenses (Cash Basis) $’s in Millions Contributions 221,459,214 80,250,000 Investment Income 424,446,359 477,382,475 Grants Paid Program Total 645,905,573 557,632,475 General and Administrative Capital Expenditures For 12 Months Ended For 12 Months Ended 272.9 Expenses 12/31/17* 12/31/16 300 231.7 Grants Paid 250 272,920,016 231,726,909 Change in Grants Payable 19,124,032 (2,954,477) Program 200 76,879,247 52,515,562 General and Administrative 22,006,106 17,386,576 150 Depreciation and Amortization 11,321,845 3,618,990 Taxes 66.5 40.4 52.6 6,802,007 9,917,523 100 50.3 24.4 39.4 Total 409,053,253 312,211,083 50

0 Net Income 236,852,320 245,421,392 2017 2016

FINANCIALS * Unaudited financial statements 42 SIMONS FOUNDATION FLATIRON FLATIRON INSTITUTE INSTITUTE SCIENTISTS SCIENTISTS

Center for Computational Astrophysics

David Alonso Justin Alsing Lauren Anderson Tjitske Starkenburg Daniel Anglés-Alcázar Amiel Sternberg Florencio Balboa Usabiaga James Stone Nicholas Battaglia Romain Teyssier Megan Bedell Stephanie Tonnesen Jeremy Magland Vasily Belokurov Francisco Villaescusa-Navarro Cara Magnabosco Jo Bovy Elijah Visbal Victor Minden Ramy Brustein Ben Wandelt Aditya Mishra Center for Computational Greg Bryan Neal Weiner Christian L. Müller Quantum Physics Matteo Cantiello Ehssan Nazockdast Kelle Cruz Center for Computational Biology Daniel Needleman Giuseppe Carleo Romeel Dave Naomi Oppenheimer David Ceperley Avishai Dekel Tarmo Äijö Cengiz Pehlevan Jing Chen Wyn Evans Joakim Andén Eftychios Pnevmatikakis Xi Chen Glennys Farrar Yanis Bahroun Daniel Podolsky Martin Claassen Scientific Computing Core Stephen Feeney Alex Barnett Aditya Rangan Antoine Georges Daniel Foreman-Mackey Meet Barot Anders Rasmussen Alexandru Georgescu Nick Carriero Charles Gammie Richard Bonneau P. Douglas Renfrew Gabriel Kotliar Justin Creveling Shy Genel Nikolai Chapochnikov David Saintillan Peter Lunts Ian Fisk Melanie Habouzit Kathleen Chen Rachel Sealfon Andrew Millis Andras Pataki Chris Hayward Xi Chen Anirvan Sengupta Olivier Parcollet Dylan Simon Lars Hernquist Dmitri “Mitya” Chklovskii Sebastian Seung Riccardo Rossi Nikos Trikoupis David Hogg Nicholas Chua Michael Shelley Angel Rubio Aaron Watters Kenta Hotokezaka Patrick Combettes Amit Singer Hao Shi Chia-Yu Hu Nick De Veaux Marina Spivak Miles Stoudenmire Jack Hughes Charles Epstein David Stein Hugo Strand Fangzhou Jiang Reza Farhadifar Mariano Tepper Steve White Chang-Goo Kim Lisa Fauci Olga Troyanskaya Manuel Zingl Yuri Levin Johannes Friedrich Shravan Veerapaneni Miao Li Julien Funk Jun Wang Yuan Li Sebastian Fürthauer Aaron Wong Christopher McKee Mariano Gabitto Witold Wysota Chiara Mingarelli Alex Genkin Wen Yan Suvodip Mukherjee Andrea Giovannucci Jian Zhou Sigurd Naess Vladimir Gligorijevic David Neufeld Kiley Graim Jerry Ostriker Leslie Greengard Lyman Alexander Page John Hayward Tsvi Piran Shidong Jiang Bill Press Jaeyoon James Jun Eliot Quataert Eva Kanso Ellianna Schwab Julia Koehler Rachel Somerville Enkeleida Lushi David Spergel FLATIRON INSTITUTE INSTITUTE FLATIRON INVESTIGATORS

44 SIMONS FOUNDATION MATHEMATICS AND MATHEMATICS AND PHYSICAL PHYSICAL SCIENCES SCIENCES SIMONS INVESTIGATORS INVESTIGATORS

Christopher Skinner Awardees Scott Aaronson Allan Sly Mina Aganagic Dam Son Vijay Balasubramanian Ian Agol Kannan Soundararajan Sam Brown Simons Collaboration on Igor Aleiner Anton Kapustin Dan Spielman Robbert Dijkgraaf Homological Mirror Symmetry Andrea Alu Ludmil Katzarkov Anatoly Spitkovsky Tony Ezome Rajeev Alur Richard Kenyon Iain Stewart Terrence Hwa Mohammed Abouzaid Sanjeev Arora Subash Khot Madhu Sudan Brian Keating Denis Auroux Ngô Bao Châu Alexei Kitaev Terence Tao Christopher Klausmeier Ron Donagi Boaz Barak Jon Kleinberg Daniel Tataru Jane Kondev Kenji Fukaya Andrei Beloborodov Kirill Korolev Shang-Hua Teng Adrian Lee Ludmil Katzarkov Simons Collaboration on the Andrei Bernevig James Lee Senthil Todadri Simon Levin Maxim Kontsevich Many Electron Problem Andrea Bertozzi Andrea Liu Chris Umans Richard Lifton Bong Lian Manjul Bhargava Madhav Mani Salil Vadhan Edward Lungu Tony Pantev Garnet Chan Dan Boneh Lisa Manning Mark Van Raamsdonk M. Cristina Marchetti Shing-Tung Yau Antoine Georges Simon Brendle Vladimir Markovic Ashvin Vishwanath Pankaj Mehta Emanuel Gull Michael Brenner James McKernan Anastasia Volovich Surjeet Rajendran Simons Collaboration on Gabriel Kotliar Garnet Chan Pankaj Mehta Aryeh Warmflash Alvaro Sanchez Special Holonomy in Evgeny Kozik Moses Charikar Maryam Mirzakhani Michael Weinstein Diaraf Seck Geometry, Analysis and Physics Olivier Parcollet Nigel Cooper Joel Moore Daniel Weissman Boris Shraiman Nikolay Prokofiev Ingrid Daubechies Andrew Mugler Horng-Tzer Yau Mukund Thattai Bobby Acharya Mark van Schilfgaarde Michael Desai Arvind Murugan Xi Yin Christopher Tully Robert Bryant Sandro Sorella Daniel Eisenstein James O’Dwyer Olga Zhaxybayeva Massimo Vergassola Simon Donaldson Guifre Vidal Alex Eskin Andrei Okounkov David Zuckerman Kalin Vetsigian Sebastian Goette Lucas Wagner Jonathan Feng Hirosi Ooguri Mark Haskins Steven White Paul François Eve Ostriker Simons Collaboration on Dominic Joyce Dominika Zgid Victor Galitski Bjorn Poonen Algorithms and Geometry David Morrison Shiwei Zhang Surya Ganguli Frans Pretorius Johannes Nordstrom Sharon Glotzer Eliot Quataert Alexandr Andoni Simon Salamon Shafi Goldwasser Leo Radzihovsky Sanjeev Arora Song Sun Ben Green Igor Rodnianski Tim Austin Steven Gubser Raphael Rouquier Mark Braverman Larry Guth Anders Sandvik Jeff Cheeger Christopher Hacon David Schwab Subash Khot Oskar Hallatschek Paul Seidel Bruce Kleiner Patrick Hayden Eva Silverstein Assaf Naor Chris Hirata Amit Singer Ran Raz Wayne Hu Oded Regev Russell Impagliazzo Michael Saks Piotr Indyk Amit Singer Shamit Kachru David Steurer Randall Kamien Avi Wigderson Marc Kamionkowski Charles Kane MPS INVESTIGATORS MPS INVESTIGATORS

46 SIMONS FOUNDATION MATHEMATICS AND MATHEMATICS AND PHYSICAL PHYSICAL SCIENCES SCIENCES INVESTIGATORS FELLOWS

Mathematics It from Qubit: Simons Collaboration on Matthew Baker Quantum Fields, Gravity and David Ben-Zvi Information Mladen Bestvina Lewis Bowen Scott Aaronson Simons Collaboration on Tobias Colding Dorit Aharonov Arithmetic Geometry, Number Panagiota Daskalopoulos Vijay Balasubramanian Simons Collaboration on the Theory, and Computation Aleksandar Donev Horacio Casini Nonperturbative Bootstrap Zeev Dvir Daniel Harlow Jennifer Balakrishnan Ezra Getzler Theoretical Physics Patrick Hayden Christopher Beem Noam Elkies Anna Gilbert Matthew Headrick Miguel Costa Brendan Hassett Florian Herzig Philip Argyres Alexei Kitaev Andrew Fitzpatrick Bjorn Poonen John Imbrie Robijn Bruinsma Juan Maldacena Thomas Hartman Andrew Sutherland Jeff Kahn Robert Caldwell Alexander Maloney Jared Kaplan John Voight Jeremy Kahn Claudio Chamon Donald Marolf Zohar Komargodski Michael Kapovich Aashish Clerk Robert Myers Joao Penedones Origins of the Universe Initiative Boris Khesin Eric D’Hoker Jonathan Oppenheim David Poland Kay Kirkpatrick Matthew Headrick John Preskill Silviu Pufu Richard Bond Nitu Kitchloo Andrew Jordan Leonard Susskind Leonardo Rastelli Claudia de Rham Alex Kontorovich Gabriel Kotliar Brian Swingle Slava Rychkov Raphael Flauger Svitlana Mayboroda Alice Quillen Tadashi Takayanagi David Simmons-Duffin Gregory Gabadadze Chikako Mese Lisa Randall Mark Van Raamsdonk Balt van Rees Anna Ijjas Tomasz Mrowka Marcus Spradlin Pedro Vieira Liam McAllister Camil Muscalu Neal Weiner Simons Collaboration on Rachel Rosen Irina Nenciu Cracking the Glass Problem Eva Silverstein Thomas Nevins Paul Steinhardt Hee Oh Ludovic Berthier Matias Zaldarriaga Julia Pevtsova Giulio Biroli Andrei Rapinchuk Patrick Charbonneau Daniel Ruberman Eric Corwin Mark Rudelson Silvio Franz Thomas Scanlon Jorge Kurchan Natasa Sesum Andrea Liu Gigliola Staffilani Lisa Manning Nicolas Templier Sidney Nagel Benedek Valkó Giorgio Parisi András Vasy David Reichman Alexander Volberg Matthieu Wyart Sijue Wu Francesco Zamponi Wei Zhang Maciej Zworski MPS INVESTIGATORS MPS INVESTIGATORS

48 SIMONS FOUNDATION SFARI SFARI INVESTIGATORS INVESTIGATORS

Kevin Eggan Evan Eichler A. Kimberley McAllister Britta Eickholt Steven McCarroll Ted Abel James Ellis Margaret McCarthy Alexej Abyzov Mayada Elsabbagh Emma Meaburn Xiaoqin Wang David Anderson Cagla Eroglu Vinod Menon Michael Wangler Dora Angelaki William Fairbrother Judith Miles Zachary Warren Michiel Basson W. Andrew Faucett Yi-Ping Hsueh Alea Mills Robert Schultz Lauren Weiss Helen Bateup Daniel Feldman Z. Josh Huang Guo-Li Ming Leah Schust Marius Wernig Mark Bear Guoping Feng Kimberly Huber Robi Mitra Ethan Scott Rachel Wevrick Kevin Bender Gordon Fishell Jun Huh Eric Morrow Nenad Sestan Tonya White Raphael Bernier Loren Frank Lilia Iakoucheva Philippe Mourrain Stephen Sheinkopf Michael Wigler Stephanie Bielas Maria Freire Ivan Iossifov Alysson Muotri Elliott Sherr Jeremy Willsey Somer Bishop Andreas Frick Fulai Jin Shrikanth Narayanan Frederick Shic Melanie Woodin Randy Blakely Harrison Gabel Elizabeth Jonas Tse Nga Ng Lisa Shulman Anthony Wynshaw-Boris Benjamin Blencowe Alfred George Emily Jones James Noonan Matthew Siegel Fei Xu Mark Blumberg Daniel Geschwind Rebecca Jones Alex Nord Steven Siegelbaum Lei Xu Susan Y. Bookheimer Jay Gibson David Julius Gaia Novarino James Sikela Shinya Yamamoto Joseph Buxbaum Charles Gilbert Kristopher Kahle Cian O’Donnell Pawan Sinha Je-Hyun Yoon Steven Chance David Ginty Raymond Kelleher Kassandra Ori-McKenney Stelios Smirnakis Haiyuan Yu Moses Chao Antonio Giraldez Albert Keung Brian O’Roak Vikaas Sohal Timothy Yu Chinfei Chen Santhosh Girirajan Alexander Kolevzon Theo Palmer Hongjun Song Chaolin Zhang Simon Chen Joseph Gleeson Genevieve Konopka Georgia Panagiotakos Matthew State Feng Zhang Gloria Choi Robin Goin-Kochel Abba Krieger Karen Parker Paul Sternberg Mingjie Zhang George Church David Goldstein Arnold Kriegstein Kevin Pelphrey Beth Stevens R. Suzanne Zukin A. Ercument Cicek Matthew Goodwin Smita Krishnaswamy Ben Philpot Garret Stuber Eli Zunder Mark Clements Alessandro Gozzi Chun-Hay Alex Kwan Michael Piper Thomas Südhof Mark Zylka Barry Connors Ann Graybiel Kenneth Kwan David Pitcher David Sulzer Edwin Cook Michael Graziano Kasper Lage Christopher Pittenger Mriganka Sur Hilary Coon Zhenglong Gu Anthony Lamantia Michael Platt James Sutcliffe Greg Cooper Adam Guastella Gary Landreth Renato Polimanti Helen Tager-Flusberg Rui Costa James F. Gusella Markita Landry Carlos Portera-Cailliau Michael Talkowski Gerald Crabtree Melissa Gymrek Maria Lehtinen Aaron Quinlan Guomei Tang Charles Craik Kurt Haas Jason Lerch Mani Ramaswami Brian Theyel Lisa Croen Michael Halassa Matthew Lerner James Rehg David Traver Colm Cunningham Antonio Hardan Bo Li Danny Reinberg Peter Tsai Mark Daly Xin He Paul Lipkin Joel Richter Gina Turrigiano Sandeep Datta Bruce Herring W. Ian Lipkin Kathryn Roeder Erik Ullian Graeme Davis David Hirsh Dan Littman John Rubenstein Flora Vaccarino Yves De Koninck Ellen Hoffman Catherine Lord Nicole Russo-Ponsaran Jeremy Veenstra-VanderWeele Bernie Devlin Mady Hornig John Lukens Stephan Sanders Pam Ventola Adriana Di Martino Liqun Luo Guillermo Sapiro Dennis Vitkup Ilan Dinstein Jeffrey Macklis Celine Saulnier Anna Docherty Robert C. Malenka Christelle Scharff Catherine Dulac Dara Manoach Devanand Manoli Gabor Marth Thomas Maynard Sarkis Mazmanian SFARI INVESTIGATORS SFARI

50 SIMONS FOUNDATION SFARI LIFE INVESTIGATORS SCIENCES INVESTIGATORS

Simons Collaboration on the Global Brain

SPARK Awardees Larry Abbott Zachary Mainen Ralph Adolphs Valerio Mante Leonard Abbeduto Misha Ahrens Markus Meister David Amaral Emre Aksay Kenneth Miller Robert Annett David Anderson J. Anthony Movshon Bridge to Independence Awards Raphael Bernier Dora Angelaki William Newsome Eric Butter Richard Axel Liam Paninski Renata Batista-Brito Laura Carpenter William Bialek Joseph Paton Graham Diering Gabriel Dichter David Brainard Pietro Perona Michael Gandal Craig Erickson Carlos Brody Bijan Pesaran Sung Han Eric Fombonne Elizabeth Buffalo Jonathan Pillow Keren Haroush Robin Goin-Kochel Matteo Carandini Xaq Pitkow Sung Eun Kwon Amanda Gulsrud E.J. Chichilnisky Alexandre Pouget Yun Li Melissa Hale Anne Churchland Jennifer Raymond Thomasz Nowakowski Suma Jacob Mark Churchland Fred Rieke Rui Peixoto Stephen Kanne Thomas Clandinin Nicole Rust Aakanksha Singhvi Catherine Lord Marlene Cohen Vanessa Ruta Holly Stessman Christa Lese Martin John Cunningham Bernardo Sabatini Jason Yi Jacob Michaelson Sandeep Datta Maneesh Sahani Cesar Ochoa-Lubinoff James DiCarlo C. Daniel Salzman Brian O’Roak Brent Doiron Elad Schneidman Opal Ousley Shaul Druckmann Krishna Shenoy Juhi Pandey Uri Eden Eero Simoncelli Karen Pierce Florian Engert Spencer Smith Joseph Piven Adrienne Fairhall Haim Sompolinsky Lisa Prock Michale Fee Michael Stryker Cordelia Robinson Ila Fiete Karel Svoboda Mustafa Sahin Loren Frank David Tank Robert Schultz Jeremy Freeman Doris Tsao Matthew Siegel Winrich Freiwald Naoshige Uchida Latha Soorya Stefano Fusi Brian Wandell Zachary Warren Surya Ganguli Xiao-Jing Wang Ericka Wodka Lisa Giocomo Ilana Witten Mark Goldman Daniel Yamins Michael Hausser Byron Yu Elizabeth Hillman Anthony Zador Mehrdad Jazayeri Manuel Zimmer Roozbeh Kiani Steven Zucker Adam Kohn Brian Lau Andrew Leifer Nuo Li Michael Long Christian Machens SFARI INVESTIGATORS SFARI

52 SIMONS FOUNDATION LIFE LIFE SCIENCES SCIENCES INVESTIGATORS INVESTIGATORS

Simons Collaboration on Ocean Processes and Ecology Project Investigators

E. Virginia Armbrust Penny Chisholm Simons Collaboration David Caron Robert DeSalle Klingenstein-Simons Fellowship on the Origins of Life Penny Chisholm Wayne Goodman Awards in Neurosciences Matthew Church Brian Hammer Donna Blackmond Edward DeLong Fritz Henn Susanne Ahmari Tanja Bosak Sonya Dyhrman Bonnie Hurwitz HHMI-Simons Faculty Scholars Benjamin Arenkiel Dieter Braun Michael Follows Eunsoo Kim Matthew Banghart David Catling Anitra Ingalls Elizabeth Kujawinski Neal Alto Jayeeta Basu Irene Chen Seth John Raghuveer Parthasarathy Thomas Bernhardt Stephen Brohawn Jason Dworkin David Karl Martin Polz Jesse Bloom Mark Churchland Woodward Fischer Debbie Lindell John Pringle Edward Boyden Richard Daneman Gregory Fournier Daniel Repeta Simons Collaboration on Principles Stephen Quake Clifford Brangwynne Benjamin de Bivort John Grotzinger Benjamin Van Mooy of Microbial Ecosystems Gene Robinson Jose Dinneny Dion Dickman Wilhelm Huck Joshua Weitz Heidi Sosik Michael Fischbach Gul Dolen Gerald Joyce Angelicque White Martin Ackermann Ramunas Stepanauskas Elizabeth Haswell Jeff Donlea Lisa Kaltenegger Jonathan Zehr Sebastian Bonhoeffer Lisa Stubbs Martin Jonikas Felice Dunn Ramanarayanan Krishnamurthy Otto Cordero David Valentine Luciano Marraffini Monica Dus Sheref Mansy Simons Collaboration on Jeff Gore William Wcislo Frederick Matsen IV Evan Feinberg Karin Öberg Computational Biogeochemical Terrence Hwa Joao Xavier Coleen Murphy Liang Feng Matthew Powner Modeling of Marine Ecosystems Naomi Levine Samara Reck-Peterson Harrison W. Gabel Didier Queloz Mary Ann Moran Simons Early Career Michael Rust Lisa Giocomo Nita Sahai E. Virginia Armbrust Victoria Orphan Investigators in Elizabeth Sattely Michael Halassa Dimitar Sasselov Jacob Bien Roman Stocker Marine Microbial Ecology Jan Skotheim Catherine Hartley Burckhard Seelig Christopher Edwards and Evolution Gurol Suel Benjamin Hayden Roger Summons Zoe Finkel Simons Collaboration Benjamin Tu Biyu He John Sutherland Michael Follows on Ocean Processes and Andrew Alverson Feng Zhang Michael Hoppa Jack Szostak Jed Fuhrman Ecology – Gradients Jake Bailey Daniel Zilberman Elaine Y. Hsiao Paula Welander Andrew Irwin Andrew Barton Mehrdad Jazayeri George Whitesides Trevor Platt E. Virginia Armbrust Erin Bertrand Andrew Kruse Brian Powell Randelle Bundy Tanja Bosak Byungkook Lim Shubha Sathyendranath Zoe Finkel Otto Cordero Conor Liston Joseph Vallino Michael Follows Anne Dekas Dengke Ma Anitra Ingalls Naomi Levine Carolyn McBride Seth John Karen Lloyd Christine Merlin Laurie Juranek Alyson Santoro Kate Meyer David Karl Frank Stewart Evan Miller Debbie Lindell Jacob Waldbauer Katherine Nagel Angelicque White Yuki Oka Jonathan Zehr Brian O’Roak Engin Ozkan Tiffany Schmidt John Tuthill Wei Xu Michael Yartsev LIFE SCIENCES INVESTIGATORS

54 SIMONS FOUNDATION LIFE SIMONS SCIENCES SOCIETY FELLOWS OF FELLOWS

Simons Collaboration on Simons Collaboration Computational Biogeochemical on the Global Brain Modeling of Marine Ecosystems Postdoctoral Fellows Fellows Senior Fellows

Sophie Aimon John Casey Boris Altshuler Katherine Cora Ames Christopher Follett Moses Chao Adam Calhoun David Heeger Junior Fellows Xiaoyin Chen Fellowships in David Hirsh Maria Dadarlat Marine Microbial Ecology Simons Fellows of the Ruth Lehmann Ruth Angus Chunyu Duan Jane Coffin Childs Memorial Fund Carol Mason Gilad Asharov Vikram Gadagkar Natalie Cohen for Medical Research John Morgan Arkarup Bandyopadhyay Anna Gillespie Keisuke Inomura J. Anthony Movshon Tobias Bartsch James Heys Chana Kranzler Brittany Belin Andrei Okounkov Sonja Billerbeck Danique Jeurissen Wei Qin David Booth Margaret Wright Michal Breker Matthew Kaufman Yunji Davenport Timothy Burbridge Aaron Koralek Simons Fellows of the Wenyan Jiang Jennifer Bussell Liang Liang Life Sciences Research Foundation Duncan Leitch Shana Caro Scott Linderman Christopher Lopez Sylvain Carpentier John Long Aakash Basu Patrick Mitchell Eric Castillo Malavika Murugan Scott Behie Joshua Modell James Dama Amy Ni Thomas Boothby Jairo Diaz Amaya Ian Oldenburg Rogier Braakman Simons Fellows of the Helen Hay Sara Fenstermacher Marino Pagan Tin Chi Solomon Chak Whitney Foundation Logan Grosenick Braden Purcell Kurt Dahlstrom Benjamin Harrop-Griffiths Evan Schaffer Romain Darnajoux Eleanore J. Clowney Keith Hawkins Sarah Davies Lihui Feng J. Colin Hill Simons Collaboration on the Robert Jinkerson Tania J. Lupoli Kohei Inayoshi Origins of Life Fellows Ricardo Laranjeiro Tomas Pluskal Dion Khodagholy Michele LeRoux Arthur Prindle Chervin Laporte Zachary Adam Hoong Chuin Lim Jeremy M. Rock Rafael Maia Clara Blättler Eric Lubeck Olena Zhulyn Bianca Jones Marlin Paul Carroll Ryan Melnyk Takashi Onikubo Claudia El Nachef Dipti Nayak Ilya Razenshteyn Teresa Ruiz Herrero Lena Pernas Carlotta Ronda Ankit Jain Benjamin Ross Aditi Sheshadri Alexandria Johnson Caroline Runyan Mijo Simunovic Sarah Rugheimer MacGregor Longfei Shu James Stafford Raghav Poudyal Michael Smith Eliran Subag Sukrit Ranjan Matthew Swaffer Xin Sun Vlada Stamenkovic David Tourigny Yi Sun Rafal Szabla Josep Vilarrasa-Blasi Li-Cheng Tsai Stephanie Valleau Christopher Whidden Michael Waskom Xingchen Wang Zheng Wu Yajun Wang Li Zeng LIFE SCIENCES INVESTIGATORS

56 SIMONS FOUNDATION OUTREACH SUPPORTED AND EDUCATION INSTITUTIONS

Adventure Scientists American Society for Cell Biology BEAM BioBus Inc. California Academy of Sciences Caveat American Academy of Arts and Sciences City University of New York Cold Spring Harbor Laboratory Cold Spring Harbor Laboratory Columbia University Medical Center DonorsChoose.org Icahn School of Medicine at Mount Sinai Gathering4Gardner Inc. Institute for Advanced Study Guerilla Science Massachusetts Institute of Technology Howard Hughes Medical Institute: The Serengeti Rules Mathematical Sciences Research Institute (MSRI) iBiology Inc.: Untitled CRISPR documentary Memorial Sloan Kettering Cancer Center IEEE Foundation: Untitled Claude Shannon documentary National Academy of Sciences Imagine Science Films Corp New York Genome Center Inc. Iridescent New York Structural Biology Center Junior Achievement of South Central PA Inc. Rockefeller University Los Angeles Performance Practice: AFTER Stony Brook Foundation Inc. Massachusetts Institute of Technology Math for America Mathematical Sciences Research Institute (MSRI): Numberphile MICRO Motherboard: The Most Unknown National Academy of Sciences National Museum of Mathematics New York Hall of Science New York Harbor Foundation, Billion Oyster Project New York Public Radio: Only Human New York Public Radio: Radiolab New York University Pioneer Works Rockaway Waterfront Alliance Inc. Rockefeller University Science Festival Foundation Science Philanthropy Alliance ScienceCounts Inc. Society for Integrative and Comparative Biology STEM Next Opportunity Fund Strategic Education Research Partnership Institute (SERP) Sundance Institute The Conversation US Inc. The Exploratorium Wave Hill Incorporated Wiki Education Foundation Wikimedia Foundation Woodrow Wilson Foundation YMCA of the USA OUTREACH AND EDUCATION

58 SIMONS FOUNDATION ADVISORY ADVISORY BOARDS BOARDS

Mathematics & Physical Sciences Spectrum Advisory Board Scientific Advisory Board Stephanie Chan Alfred Aho Zwoop Columbia University SFARI Scientific Advisory Board Michael E. Goldberg Charles Epstein David Lewis Columbia University University of Pennsylvania University of Pittsburgh Laura Helmuth Nicholas M. Katz Richard Lifton Washington Post Princeton University Rockefeller University Quanta Advisory Board Robin Marantz Henig Alfred Mueller Eric Nestler The New York Times Magazine Laura Chang Columbia University Icahn School of Medicine at Mount Sinai The New York Times Ivan Oransky Ramesh Narayan Martin Raff New York University Raissa D’Souza Harvard University University College London University of California, Davis Aviv Regev Christos H. Papadimitriou Arnon Rosenthal Massachusetts Institute of Technology David J. Gross Columbia University Alector LLC Kavli Institute for Theoretical Physics David Sassoon Jill Pipher Carla Shatz InsideClimate News Hopi E. Hoekstra Brown University Stanford University Harvard University Will Talbot Karin Rabe Elizabeth Spelke Stanford University Alex Kontorovich Rutgers, The State University of New Jersey Harvard University Rutgers, The State University of New Jersey

Srinivasa Varadhan Huntington F. Willard Howard Schneider New York University University of Chicago Stony Brook University School of Journalism

Margaret H. Wright Steven Strogatz New York University Cornell University

Leslie B. Vosshall Rockefeller University ADVISORY BOARDS ADVISORY

60 SIMONS FOUNDATION ADVISORY BOARD BOARDS OF DIRECTORS

SPARK Advisory Board

David Eisenbud, Ph.D. Paul S. Appelbaum Director, Mathematical Sciences Research Institute Columbia University

Gerald D. Fischbach, M.D. Science Sandbox Advisory Board Guy Calvert Distinguished Scientist and Fellow, Simons Foundation Dup15q Alliance Bruce Alberts Margaret A. Hamburg, M.D. University of California, San Francisco Monica Coenraads President, American Association for the Advancement of Science Rett Syndrome Research Trust Alan Alda Peter Littlewood, Ph.D. Alan Alda Center for Communicating Science, Daniel Geschwind University of Chicago Stony Brook University University of California, Los Angeles

William H. Press, Ph.D. Majora Carter Omar Khwaja University of Texas at Austin MCG Consulting Roche Pharmaceutical Research StartUp Box and Early Development Mark Silber, J.D., M.B.A. Executive Vice President and Chief Financial Officer, Renaissance Technologies Kishore Hari Ami Klin Emory University School of Medicine James H. Simons, Ph.D. Werner Herzog Chair, Simons Foundation Paul Lipkin Miranda July Kennedy Krieger Institute Marilyn H. Simons, Ph.D. President, Simons Foundation Robert Lue Sandy Magaña Harvard University University of Texas at Austin Shirley M. Tilghman, Ph.D. Princeton University Vikki Spruill Heather C. Mefford University of Washington

Steven Shore Adelphi University

Alison Singer Autism Science Foundation

Bridget A. Taylor Alpine Learning Group ADVISORY BOARDS ADVISORY

62 SIMONS FOUNDATION SIMONS FOUNDATION SIMONS FOUNDATION STAFF STAFF

Mani Cavalieri Yuan Li Nadine Celestin Thomas Lin Amiel Sternberg Jordana Cepelewicz Paula Lukats James Stewart Ahmad Abbad Nikolai Chapochnikov Vladimir Gligorijevic Peter Lunts Colleen Stock Ilona Abramova Jing Chen Katie Goodwin Cindy Rampersad-Phillips Alice Luo Clayton Benjy Stokar John Acampado Kathleen Chen Kiley Graim Rishi Rana Enkeleida Lushi Miles Stoudenmire Andrea Ace Xi Chen Anastasia Greenebaum Anders Rasmussen Jeremy Magland Hugo Strand Stephanie Adika Xi Chen Leslie Greengard Lucy Reading-Ikkanda Cara Magnabosco Thomas Sumner Maria Adler Wu-bin Chin Marion Greenup Louis Reichardt Jennifer Maimone-Medwick Emily Tan Leyla Ahari Dmitri “Mitya” Chklovskii Michael Grey Matthew Reiser Malcolm Mallardi Rebecca Tancredi Tarmo Äijö Andrew Choi Ann Griswold P. Douglas Renfrew Apoorva Mandavilli Mariano Tepper Justin Alsing Salim Chowdhury Luke Grosvenor John Rennie Richard Marini Allegra Thomas Alpha Amatya Nicholas Chua Brigitta Gundersen Woody Richards Michelle Matias Suraj Tiwari Joakim Andén Martin Claassen Patrick Gunn Christopher Rigby Richard McFarland Jennifer Tjernagel Lauren Anderson Carleen Clarke Melanie Habouzit Samantha Riviello Andrew Millis Stephanie Tonnesen Aireli Angel-Ramos Secoha Cooper Fang Han Mark Roberts Victor Minden John Tracey Daniel Anglés-Alcázar Justin Creveling Sunita Hansraj Beverly Robertson Chiara Mingarelli Nikos Trikoupis Caleb Arnold Neha Dandu Dominique Harrison Mariah Roda Jillian Minogue Olga Troyanskaya Irina Astrovskaya Amy Daniels Kevin Hartnett Edgar Rodriguez Emily Miraldi Yuri Tschinkel Kate Augenblick James Davidson Marcus Haugen Jowy Romano Aditya Mishra Dawn Tucker Yanis Bahroun Nick De Veaux Christopher Hayward Ricardo Rossi Katie Moisse Hope Vanderberg Florencio Balboa Usabiaga Donna DeJesus-Ortiz John Hayward Anthony Roux Daniel Mortensen Benjamin VanderSluis Shareen Bamberg Jocelyn Dorszynski Mary Kate Hennelly Elizabeth Roy Michael Moyer Jan Varghese Alex Barnett Shaun Dubreuil Deborah Hertz Angel Rubio Elizabeth Mrozinska Brianna Vernoia Meet Barot Reza Farhadifar David Hogg Nick Sanghvi Suvodip Mukherjee Francisco Villaescusa-Navarro Agnes Barszcz Meghan Fazzi Jessica Holthouser Diane Sarria Christian L. Müller Elijah Visbal Asif Bashar Stephen Feeney Rebecca Horne Kathleen Savarese Megan Muneeb Natalia Volfovsky Nicholas Battaglia Pamela Feliciano Kenta Hotokezaka Alyssa Picchini Schaffer Vincent Myers Karen Walton-Bowen Megan Bedell Gerald D. Fischbach Chia-Yu Hu Tara Schoenfeld Sigurd Naess Ben Wandelt Jessica Bee Ian Fisk John Jagard Kim Scobie Layla Naficy Jun Wang Marta Benedetti Chris Fleisch Marian Jakubiak Rachel Sealfon Ehssan Nazockdast Paul Wang Serena Bianchi Nina Fleiss Bill Jensen Rebecca Sesny David Nelson Aaron Watters Lawrence Bianco Neil Flood Lydia Jung Swapnil Shah Camille Norrell Patricia Weisenfeld Aaron Biscombe Patrick Flood Rachel Jurd Alexandra Shaheen Eirene O’Connor James Whalley Jill Blackford Steven Ford Tim Kane Michael Shelley Sean O’Connor Casey White Lehman Alexandra Bolter Dan Foreman-Mackey Jeanette Kazmierczak Jesse Sherwood Debra Olchick Ingrid Wickelgren Rich Bonneau Johannes Friedrich Marlow Kee Hao Shi Naomi Oppenheimer Ursula Wing Greg Boustead Tammi Fumberi Matthew Kent Melanie Shiree Kristin Ozelli Natalie Wolchover Michelle Bradshaw Julien Funk Deborah Kenyon Olena Shmahalo Joanna Pacholarz Aaron Wong Libby Brooks Hannah Furfaro Chang-Goo Kim Dylan Simon Alan Packer Annie Wong Jennylyn Brown Sebastian Fürthauer Patricia Kim Chaim Singer Olivier Parcollet Paul Wong Gregory Bryan Gregory Gabadadze Emily Klein Emily Singer Andras Pataki Jessica Wright Martin Butler Mariano Gabitto Julia Koehler Kori Smith Danielle Patch Philip Yam Claire Cameron Annaliese Gaeta Michael Kranz Lee Anne Green Snyder Balmes Pavlov Wen Yan Matteo Cantiello Swami Ganesan Abe Lackman Rachel Somerville Cengiz Pehlevan Shirley Ying Jacob Cappell Valerie Gar Alex Lash Julia Sommer Nicole Phillips Cindy Young Giuseppe Carleo Jennifer Garcia Noah Lawson David Spergel Olivia Pinney Michelle Yun Nick Carriero Alexandra Geldmacher Seran Lee-Johnson John Spiro Eftychios Pnevmatikakis Hana Zaydens Lindsey Cartner Shy Genel Monika Lenard Marina Spivak Daniel Podolsky Nicholette Zeliadt Antoine Georges MaryKate Levi Christopher Sprinz Christina Pullano Jian Zhou Alexandru Georgescu Yuri Levin Tjitske Starkenburg Allison Rains Manuel Zingl Andrea Giovannucci Miao Li David Stein Alexandra Stephens STAFF

64 SIMONS FOUNDATION 160 Fifth Avenue New York, NY 10010 Tel. 646 654 0066 simonsfoundation.org sfari.org